Thread milling - How to get IJK code with non default planes?

[guilloche]

Greetings,

I am using the generic Haas mill posts.

When I thread mill, sometimes mcam 2017 puts out code like this:

N730 T6 M6
N740 G0 G90 G55 X0. Y0. A0. S6000 M3
N750 G43 H6 Z.1
N760 M8
N770 G1 Z-.27 F40.
N780 X.0887 F24.
N790 G41 D6 X0. Y-.3415
N800 G3 X.4415 Y0. I.0887 J.3415
N810 Z-.2283 I-.4415 J0.
N820 Z-.1867 I-.4415 J0.
N830 Z-.145 I-.4415 J0.
N840 Z-.1033 I-.4415 J0.
N850 Z-.0617 I-.4415 J0.
N860 G1 X.4396 Y.0413 Z-.0611
N870 X.4338 Y.0823 Z-.0605
N880 X.4241 Y.1226 Z-.0598
N890 X.4108 Y.1618 Z-.0592
N900 X.3939 Y.1995 Z-.0586
N910 X.3734 Y.2355 Z-.058
N920 X.3497 Y.2694 Z-.0573
N930 X.323 Y.301 Z-.0567
N940 X.2934 Y.3299 Z-.0561
N950 X.2612 Y.356 Z-.0555
N960 X.2267 Y.3788 Z-.0549
N970 X.1902 Y.3984 Z-.0542
N980 X.1521 Y.4145 Z-.0536
N990 X.1126 Y.4269 Z-.053
N1000 X.0722 Y.4356 Z-.0524
N1010 X.0311 Y.4404 Z-.0517
N1020 X-.0103 Y.4414 Z-.0511
N1030 X-.0516 Y.4385 Z-.0505......

The resulting thread looks faceted, as though "linearize helixes" were enabled, though it is not.

Sometimes it will put out code with I, J, K instead, and these threads come out perfectly smooth:

N1910 T6 M6
N1920 G0 G90 G54 X.0119 Y.9853 S6000 M3
N1930 G43 H6 Z.1
N1940 M8
N1950 G1 Z-.02 F40.
N1960 G41 D6 X0. Y.886 F24.
N1970 G2 X.786 Y0. I-.1064 J-.886
N1980 Z-.0617 I-.786 J0.
N1990 Z-.1033 I-.786 J0.
N2000 Z-.145 I-.786 J0.
N2010 Z-.1867 I-.786 J0.
N2020 Z-.2283 I-.786 J0.
N2030 Z-.27 I-.786 J0.
N2040 Z-.3117 I-.786 J0.
N2050 X.7701 Y.1573 Z-.352 I-.786 J0.
N2060 X.1773 Y-.8681 I-.8743 J-.1786
N2070 G1 G40 X.2088 Y-.963
N2080 G0 Z.1
N2090 M5
N2100 G91 G28 Z0. M9

The good IJK code is Op 1, G54 with the default work offset and top plane. The non IJK code is Op 2, the back side of the part, G55 with a custom plane, work offset #1.

In machine control settings, helix support is set to all planes.

How to get the good IJK code with non default planes?

[JParis]

Can you share a z2g file?

[guilloche]

Of what?

[JParis]

your file that's giving you linearized code

 

z2g format is best to make certain all of your settings are with the file

[guilloche]

Some kind of a bug I guess. I changed the op1 toolpaths parameters to match op 2, (from OD thread to ID, different depth, different thread diameter) but didn't change the plane, and it put out the bad code. I then changed all parameters one by one to find the cause. It seemed related to the depth. I then changed the depth over and over, incrementally moving towards the depth of the original op 1 toolpath to see where the change occurred. I got within .003" of the op 1 depth and then it put out IJK code, so then I put in a higher value again expecting to see bad code, but now no matter what depth I put in, or if I changed the plane, or if I select the point that the op2 toolpath is based on, or whatever, it puts out the good code. So I just copy pasted that toolpath into op2 and my problem is solved....