Using U axis without X

Hi,

I was running my machine yesterday (dual-table router) and since the X table is having some issues we decided to modify the program to run on the U axis table (find and replace all X with U in the Gcode). I found the following issues when not using the X axis:

1. The preview toolpath adjusts to the position of the X axis regardless of the fact that there are no Xs in the program. You can't run the program unless the X axis thinks it's in the right position despite the U offsets being set correctly.
2. The U axis ignored all feed speed commands and moved at the rapid speed at all times. The machine would fly down the U axis and then crawl across the Y axis.

Is this normal behavior for AXIS? Are there tweaks I could make to have both tables usable separately?

Thanks in advance!

Rob

LinuxCNC doesn't really know how U, V and W map into cartesian space (nor, for that matter, A, B and C)
So, path blending and velocity limits are not enforced (though feed rates, at least, should be honoured)

This isn't a simple fix, because it isn't simple to define what the extra 6 axes mean, for machines that have them, it will tend to be different from one machine to the next.

In your situation, a better idea would be to copy the INI file to a new name, and have it reference a different HAL file that passes X-commands to the U-motor. You can then choose the config you want from the config picker. The configs can share all other files.

That limitation makes sense and so does your solution. As for the feed speed I could make it slow down with the feed override but that slows every axis.. The U axis was definitely traveling 2-3 times faster than it should have been.

As an aside, I'm trying to figure out RTCP and I can't find the spot where I can enter the pivot length of the machine in the HAL or INI files. I'm hoping I don't have to compile the 5axiskins every time I need to adjust that parameter during machine testing...

Thanks!

Some nice RTCP configs are included in 2.8/Master, with some good docs.

linuxcnc.org/docs/devel/html/motion/5-axis-kinematics.html

You don't have to understand the matrix maths, that hard stuff is already in the kins modules,

github.com/LinuxCNC/linuxcnc/tree/9d6f9e...table-rotary-tilting

I think the standard 5axiskins is actually all I need for the machine but it's just that one parameter (pivot length) that I can't seem to find in any of the documents or the sample configs. Now maybe that's just because I don't know what I'm looking for...

Any help would be really appreciated.

Thanks!

Rob

Sorry, I don't know much about configuring 5 axis kins, but as to your feed rate problems with U. When the new trajectory planner first came out, there were a lot of bugs that needed to be shaken out of the way U,V,W behaved. (I helped find a few with our XYZW machine). Maybe there are still a few crawling around in there. Perhaps it would be worthwhile to try experimenting with 2.6 a little, for a basis for comparison, and see if your U problems are more well behaved.

5axiskins creates a HAL pin called "5axiskins.pivot-length" that can be connected the the tool-length HAL pins.
I think it works best if connected to the W-offset and if tool lengths are put in the W column. The tool length probably needs to be the distance from pivot to tool tip.

I have never tried to use it....

Thanks for your help!

I think that's what I need. I'm assuming with a little HAL connection magic I could set a fixed pivot length that's the distance from the pivot to the tip of the spindle and then add to that the W offset for each tool? I'm still learning what this stuff is capable of... and it seems like pretty much anything and everything.

CanadianNC wrote: I'm assuming with a littgle HAL connection magic I could set a fixed pivot length that's the distance from the pivot to the tip of the spindle and then add to that the W offset for each tool?

Yes, that should work.
It all depends on how you measure the tool length. Unfortunately I don't think that normal touch-off methods will work, but it's worth experimenting to find what does happen.