In machine simulation behaviour different from g-code

[Edit3D.studio]

Hi, I'm working on a VMC with a 4th axis. The issue is that the movements in the simulator are different from the G-code. I attached a video showing the behavior—notice the Z-axis. In the simulator, it stays the same, but in the G-code, it behaves differently. Why is that?

[programming2C78B]

You attached a post, but no gcode.

First guess is that the simulator doesn't show all transition moves, and the simulator isnt running off the gcode itself. 

[Edit3D.studio]

I setup another simulation with a simple toolpath, of a cylinder diameter 50mm. It seems both Y and Z wrongly moves of the haf of the diameter, 25mm.

I also built the machine again from ground up but with the same result, I'm lost now

PS: .ngc files are not supported to upload? anyway I uploaded as .txt

[a.laasW8M6T]

Hi

You just need to turn off TCP in the machine definition

[Edit3D.studio]

Thanks, this solved the second part of the issue, the weird rotation at the end. Anyway there is still a mismatch between machine movement and the g-code, it's visible in the video, the Z an Y floats between 0 and 25 (that is the cylinder radius in this case) but the machine is visibly still at Y0 and Z0 

[a.laasW8M6T]

Hi

Did you edit the machine in the setup or in your machine library?

When you add a machine to a document it is no longer linked to the Library machine.

Make sure you turn TCP off in the machine in your setup and try again.

I have tested with the Brother speedio post and get correct results, I don't have your post to confirm

[Edit3D.studio]

Thank you again for the help, my post processor is on the first post of this thread if you have time to check for it. In simulation (info) there are 2 locations to see where the machine is, I see the issue in "position" part (info) the "machine" position is correct

[a.laasW8M6T]

Hi

Sorry I missed that the post was there in the 1st post

You can ignore the values in the position readout, they are only correct in the original coordinate system that's not rotating

[Edit3D.studio]

Oh okay,  I tought these should be reflect exhactly the G-code when in TCP mode.. but is'nt it worrying that the simulation does't really reflect the g-code simulated as in the second video? Luckly I noticed it while cutting air..

by the time we are here, To avoid the final rewind of the A axis can this do the trick? 

Use Incremental Commands 

• Use H Codes: In your post-processing, switch from absolute (C) angular commands to incremental (H) commands, which can help manage values more effectively for multi-axis operations.

[a.laasW8M6T]

Hi

I've not tried converting a post to incremental before

You can look at the post guide in Section 5.1.9

https://cam.autodesk.com/posts/posts/guides/Post%20Processor%20Training%20Guide.pdf

If there is a way to make the axis Cyclic in your control that would be much better though