Help! discrepancy between backplot & simulation

[BradyCNC]

Just trying a standard raster, my backlot shows a smooth program but my simulation shows crazy moves. Who can i trust?

[#Rekd™]

Honestly neither (Unless you have licensed the Machsim module and have a Kinematic aware post) ....G-Code simulation or verification is the most accurate.

 

[BradyCNC]

Would something like CIMCO be sufficient for verification?

[Jake L]

The verification looks like you're trying to design a roller coaster. 

Never seen that before, any chance you can share the file?

[crazy^millman]

Refer to my arc error posting change NC tolerance to a tighter tolerance in the Machine Control Definition. That might help it has helped 10 different projects with the same exact issue recently. I am doing a project in Metric and using the post to convert to Inch. With only .0000" and .000mm I was getting all kind of arc errors on Siemens 840D control. I went down to .000000" and .00000mm and they went away except for mapped planes. On several HAAS projects I was getting the same errors showing up in Vericut. I changed them to .00000" and .0000mm and they went away there. Matsuura using a Mastercam Post was doing the same thing in Vericut I used the tighter tolerance and those issue went away there also. CAMPLETE figured out NCI was prone to truncation and rounding errors because the arcs are calculated wrong when passed from NCI to MP.Dll, back to NCI then to the Post when done in Mastercam. They wrote their own logic to fix this possible arc error so they are ahead of the curve. ICAM does the same thing when they post code they check for this rounding error and prevent it, Mastercam Posts cannot do this because they are getting truncated NCI data. I have been seeing errors like this ever since X came out. Simple Math test is take a 100" Sphere and use a 1" ball endmill to cut using flowline. Run the best filters Mastercam has and post code. It will tell you everything you need to know how bad the math is in Mastercam. One G3 or G2 line should only need to be posted when allow 360 arcs is used in the MCD. 15 years I have complained and it has fallen on deaf ears. I have converted the backplotted code to arcs for every version and then repost using that backplotted geometry using no filters and it will work. Why is that? Why can't one of the core toolpaths in Mastercam work like it did in V9 20+ years ago?

ARC error

[Jobnt]

5 minutes ago, crazy^millman said:

Why can't one of the core toolpaths in Mastercam work like it did in V9 20+ years ago?

Preach it, brother!!