cancel
Showing results for 
Search instead for 
Did you mean: 

Inverse time problem

Creator
Creator


 I have a postprocessor set up for inverse time and occasionally it puts out a G94 with an angle in the line (see example code below).
I thought if there is an angle in the move it should be G93. Anyone know if this is correct behavior and if not how to stop the post from putting a G94 out on a line with rotary motion. I have tried changing tolerance etc. in post but to no avail.


G94 X.1931 Y-1.1888 Z7.2235 F8.
X.195 Y-1.1643 Z7.2221
X.1951 Y-1.1633
G93 X.1952 Y-1.1604 Z7.2222 A-90.499 C270.323 F2657.298
X.1953 Y-1.1574 C270.322 F2726.928
X.1954 Y-1.1545 Z7.2223 A-90.5 C270.321 F2717.042
X.1955 Y-1.1515 Z7.2224 C270.32 F2707.217
G94 X.1956 Y-1.1486 Z7.2225 C270.319 F8.
X.1957 Y-1.1456 Z7.2226 A-90.501 C270.318
X.1958 Y-1.1426 Z7.2227 C270.317
X.196 Y-1.1396
X.1961 Y-1.1366 Z7.2228 C270.316

Rick D

Win7 64 bit w/NX10.0.3 mp11
Vericut 8.0

13 REPLIES

Re: Inverse time problem

Esteemed Contributor
Esteemed Contributor

Guessing...

Add some code to output mom_prev_out_angle_pos and mom_out_angle_pos 

Or more correctly, the result of

[EQ_is_equal $mom_prev_out_angle_pos(n) $mom_out_angle_pos(n)[

(n = 0, 1)

 

It could be it compares mom_pos instead:

[EQ_is_equal $mom_prev_pos(n) $mom_pos(n)[

(n = 3, 4)

 

It could be EQ_is_equal reports they are the same when they really aren't...

(although a .001 difference should be "different")

 

I'm not sure where in code (or NX) it checks this.

Ken Akerboom Sr CAx Systems Engr, Moog, Inc.
Production: NX10.0.3.5 MP5 + patch/TC11.2
I'd rather be e-steemed than e-diseaseled


Re: Inverse time problem

Legend
Legend

Hi,

Do you define the global variable mom_sys_contour_feed_mode?

global mom_sys_contour_feed_mode

set mom_sys_contour_feed_mode(LINEAR)        "FRN"
set mom_sys_contour_feed_mode(ROTARY)        "FRN"
set mom_sys_contour_feed_mode(LINEAR_ROTARY) "FRN"

Re: Inverse time problem

Creator
Creator

I tried your code even though I have FRN set in feedrates, however if I use the:

 

set mom_sys_contour_feed_mode(LINEAR)        "FRN"

 

Then all my moves included plain 3 axis moves come out G93 with feedrate on every line.

 

It may be as stated in above reply that the angular difference is so small.

Re: Inverse time problem

Valued Contributor
Valued Contributor

Hi Rick,

 

In postbuilder under the Machine tab > Fourth Axis and Fifth Axis. Set the "Rotary Motion Resolution" to a smaller number. This will set the variables  "mom_kin_4th_axis_min_incr   and mom_kin_5th_axis_min_incr" in the main TCL file. These variables are used in the "ugpost_base.tcl" file to set the feed type in the lines

 

  set tol_rot4 $mom_kin_4th_axis_min_incr

  set tol_rot5 $mom_kin_5th_axis_min_incr

 

  if { ![EQ_is_equal_tol $mom_pos(3) $mom_prev_pos(3) $tol_rot4] || \
   ![EQ_is_equal_tol $mom_pos(4) $mom_prev_pos(4) $tol_rot5] } { set isrotary_move 1 }

 

This is the only place I could find that used those variables so it should be safe to change them but as usual test,test,test.

 

HTH Eric

Re: Inverse time problem

Esteemed Contributor
Esteemed Contributor

 

 

  set tol_rot4 $mom_kin_4th_axis_min_incr

  set tol_rot5 $mom_kin_5th_axis_min_incr

 

  if { ![EQ_is_equal_tol $mom_pos(3) $mom_prev_pos(3) $tol_rot4] || \
   ![EQ_is_equal_tol $mom_pos(4) $mom_prev_pos(4) $tol_rot5] } { set isrotary_move 1 }

 


NOTE:

I've found (in proc SUPER_FEED_MODE_SET) cases where mom*_pos were "not equal" but mom*_out_angle_pos were.

So you might try changing the above to mom_out_angle_pos / mom_prev_out_angle_pos.

Ken Akerboom Sr CAx Systems Engr, Moog, Inc.
Production: NX10.0.3.5 MP5 + patch/TC11.2
I'd rather be e-steemed than e-diseaseled


Re: Inverse time problem

Creator
Creator
That seems to solve the problem. I set the resolution from .001 to .0001 and not getting G94 with angle move.
Thank you.

Re: Inverse time problem

Creator
Creator
I set the rotary resolution from .001 to .0001 and added the code with mom_out_angle_pos and seems to have solved the problem putting out a G94 with an angle move.
Thank you.

Re: Inverse time problem

Valued Contributor
Valued Contributor

Sorry just seen this thread....

 

Derek Hart/Mark Rief.

 

This exact question has come up in the past, on the oldie goodie CAM BBS, & I believe it was answered by Akerboom...??  I remembered because I took notes & have used the solution ever since...

 

Having said that, is there a way to pin the old CAM BBS, right below the "Read this first - NX Community Guidelines"? That way it is accessible to everyone on here… Link below….

 

https://community.plm.automation.siemens.com/t5/NX-CAM/bd-p/Customer-NXManufacturing-CAM

 

Kal.

NX 9.0.3.4 MP12
PB 9.0.3.2
Vericut7.3.4

Re: Inverse time problem


Kal wrote:

Sorry just seen this thread....

 

Derek Hart/Mark Rief.

 

This exact question has come up in the past, on the oldie goodie CAM BBS, & I believe it was answered by Akerboom...??  I remembered because I took notes & have used the solution ever since...

 

Having said that, is there a way to pin the old CAM BBS, right below the "Read this first - NX Community Guidelines"? That way it is accessible to everyone on here… Link below….

 

https://community.plm.automation.siemens.com/t5/NX-CAM/bd-p/Customer-NXManufacturing-CAM

 

Kal.


@Kal we don't intend to list the archive for browsing, since access is limited to customers. However, it is included in search results. 

Mark Rief
Retired Siemens

Learn online





Solution Information