I've been having some dramatic bizareness on Mach 3 Beta 1.90.36 that wasn't present on 1.90.30 - with some arcs (especially the second G03 in a row), the curve is drawn correctly in the preview (Program Run screen), but the tool doesn't follow the path drawn when making the part (though it knows what it's actually doing - the erroneous path it follows is traced as it actually moved, even though it is at odds with what the intial pre-cut rendering produced). I was careless and didn't do a dry run with floral foam, so it zorched the actual part the first time. I now have a whole library of my normal part code (that worked under Mach 2) that cannot run under Mach 3. This is as bad as the comments-what are comments? bug. I'm rolling back to Mach2 for the present.

Help!

Here is a code chunk that does it (look for the comment midway through):

#40=1   ( UL,UR,LL,LR )
G40               
G17             
G49             
G80           
G90           

#04=[3/16]       
#05=[#04/2    ]
#06=[3/25.4   ]
#08=20           
#09=0.2         
 
#52=[ 1/25.4 ]       
#53=#52
#55=1             
#58=[[#55*2]-1]   

F #08               
G01 X 0 Y 0 Z #09   
     
#12=FUP[#53/#06]         
#13=[#53/#12]           
G01 Z 0     


G40       
G01 Z #09       
M3
G01 X [-7.9193+[#05*#58]] Y -0.125
G[41+#55] P #05
G01 X -7.9193 Y -0.125
F [#08/2]               
G01 Z [-1*#13]   

F #08         
G01 X -7.9193 Y 2.1654
G02 X -7.7618 Y 2.3228 R .1575
G01 X -7.2638 Y 2.3228
G02 X -7.1063 Y 2.1654 R .1575
G01 X -7.1063 Y 2.1024
G03 X -6.9488 Y 1.9449 R .1575
G01 X -6.5157 Y 1.9449
G03 X -6.3583 Y 2.1024 R .1575
G01 X -6.3583 Y 2.1654
G02 X -6.2008 Y 2.3228 R .1575
G01 X -5.7677 Y 2.3228
G02 X -5.6102 Y 2.1654 R .1575
G01 X -5.6102 Y 2.1024
G03 X -5.4528 Y 1.9449 R .1575
G01 X -5.0197 Y 1.9449
G03 X -4.8622 Y 2.1024 R .1575
G01 X -4.8622 Y 2.1654
G02 X -4.7047 Y 2.3228 R .1575
G01 X -4.2717 Y 2.3228
G02 X -4.1142 Y 2.1654 R .1575
G01 X -4.1142 Y 2.1024
( this next line renders correctly before execution, but the tool deviates here)
G03 X -3.9566 Y 1.9450 R .1575
G01 X -3.5236 Y 1.9449
G03 X -3.3661 Y 2.1024 R .1575
G01 X -3.3661 Y 3.3465
G02 X -3.2087 Y 3.5039 R .1575
G01 X -2.7756 Y 3.5039
G02 X -2.6181 Y 3.3465 R .1575
G01 X -2.6181 Y 2.1024
G03 X -2.4606 Y 1.9449 R .1575
G01 X -2.0276 Y 1.9449
G03 X -1.8701 Y 2.1024 R .1575
G01 X -1.8701 Y 2.1654
G02 X -1.7126 Y 2.3228 R .1575
G01 X -1.2795 Y 2.3228
G02 X -1.122 Y 2.1654 R .1575
G01 X -1.122 Y 2.1024
G03 X -.9646 Y 1.9449 R .1575
G01 X -.5315 Y 1.9449
G03 X -.374 Y 2.1024 R .1575
G01 X -.374 Y 2.1654
G02 X -.2165 Y 2.3228 R .1575
G01 X .2165 Y 2.3228
G40     
G01 Z #09         
     
G01 X 0 Y 0 Z #09       


M30       

Hi,

I have checked your code and can find nothing wrong with it,  I think this is one for Art but he is at the CNC-workshop meeting.

Graham.

Thanks! Nothing beats a second set of eyes willing to look code over. I hate to bother Art with noise. -- Rob

Thanks!. Art has addressed this in 1.90.38, looks good.