Weird behaviour on G02 / 1.2115

[Pulcik]

Hi guys,

I experienced a strange issue today and I am trying to figure it out.
First of all, what's changed in my setup, is the version. I upgraded to 2115, from 2110 (I think).
Everything looks same, after upgrade and seems to work (I use my own screen / under development).
But when I milled first job, I see the Z axis was weirdly slightly off the program. After little bit of examining I found, that Z axis goes in to material "slower" than it suppose to go. The weird part of it is, that UCCNC know about that as it is driving the Z axis like that (visible on DRO).

Here is short clip of that phenomenon :
[youtube]https://youtube.com/shorts/WV51-UIGqWg[/youtube]

In G02 the Z axis goes to the Z coordinate much slower than expected (even by toolpath). It is not finishing the Z move in that same line execution and extending it to another line. It just goes with its own speed and when it reaches the coordinate, then it reaches.
On the video, it happens when the line starting with X78.5 is executed. Then the Z axis slow down with feed from the previous move (almost same one), and when next line is executed, it is not where it supposed to be, by toolpath.
After little bit of playing with that phenomenon it seems, that it happens only with G02 moves with Z coordinate (ramping, boreing). And I think, it happens only with the last G02 with Z coordinate.
It means, if there are 4 lines of G02 X Y Z in a row, only with the last one this happens. The previous ones are executed correctly.
First I thought, that G64 is causing this. But it happens even with G61. The only time, when this doesn't happen is when the code is executed line by line. Than the moves are exact.
Is there a possibilities, that when UCCNC doesn't see Z coordinate ahead he decided that he have more time to go down??

I am thinking to reinstall the program, just in case something goes wrong during installation. It wasn't fresh installation, and I am experiencing some issues with G31 probe moves too. But that is something I will debug tomorrow.

Any idea what can cause this? Am I missing something? Is there a new feature I don't know about?

[Pulcik]

Forgot to mention, that I am using UC100.

As I read the forum, I will try downgrade to 2114, as I am using UCR201, and there seems to be a bug with its plugin and probing. Might be the g31 issues I have.

[cncdrive]

Please post your profile file and an example g-code file in which this issue happens, so we could try to reproduce it.

[Pulcik]

Here is the code and profile.
I try it on my demo instance at home, and it is exactly the same issue.
I did some more testing in demo mode.
When I add the Z coordinate to the following line (the same as previous line = stay on this Z cordinate), nothing change and it is still going down by its own feed and extending to other line execution. But when I slightly change this added Z coordinate, the whole problem move by one line. It seems, it is an issue of the last NEW Z coordinate.

[Pulcik]

I downgraded to 2114 and all is good now.
Both G02 moves, and probing are working as expected in 2114. About probing you already know, as I saw in another post.

So, it seems to be another bug in 2115.

[cncdrive]

Dezsoe, please test this and let me know the results.

[dezsoe]

It's also wrong in 1.2116. I tried without G43, but it was the same. Also, the Z DTG shows strange values while Z moves in G02.

(Don't search for 1.2116: it hasn't been released yet.)

[Pulcik]

Understood

[cncdrive]

We debugged and fixed this issue now, it will be OK in the next release.
I'm not sure in which version this bug got into the software, but we will investigate it.

[Battwell]

is this in the "probefix 115 version"? if so any chance of a g2fix version?