Advanced Search

Wow, ambitious, looking forward to seeing your results.

Thankyou for your contributions.

New TP2 shall have 9 axis look ahead, we are working on a different base source, i can't share a work in progress, I will be porting later, still working out feed override spikes at the moment, and thinking about remaking the TP base completely.

For multiaxis, I started also on porting and improving G12.1 (switchkins from rushabhGH), and I'm juggling with an idea to have tool length be something adjacent to a virtual axis, and if those are in place, have the proper axis limits be derived from joints + current kins + current tool

Time will tell...

sounds like graphics drivers likely loaded just fine

Have you made any edits to the file /etc/lightdm/lightdm.conf yet?

Something I think to try, before running startx:
sudo systemctl status display-manager --no-pager -l
and
systemctl status lightdm --no-pager -l
If you can, snap a good photo of each command's result and post here that will save you lots of typing.

right .... oops

5280 wrote:

Kernel driver in use 1915
Kernel modules: 1915

I think got it all typed right.

Almost, it is i915.

The 2nd reply from the program in terminal is the same.

the reply is :

00:02:0 VGA compatible controller (0300): Intel Corporation Haswell-ULT Integrated Graphics Controller (8086:0a16) (rev 09)
Subsystem: Dell Device (1028:05ec)
Kernel driver in use 1915
Kernel modules: 1915

I think got it all typed right.

Again THANK YOU for all the help.

My GladeVCP panel fails to launch intermittently. The only errors that are found in the report are:
GladeVCP.QTVCP.QT_ISTAT][^[[33mWARNING^[[0m]  INI Parsing Error, No MIN_SPINDLE_0_OVERRIDE Entry in DISPLAY, Using: 0.5 (qt_istat.py:532)
**** GLADE VCP ERROR:    With xml file: crosshair_panel.glade : 'gi.repository.Gtk' object has no attribute 'glade'

and:

crosshair_panel.hal:7: Pin 'crosshair_panel.crosshair-ctl-active' does not exist

The second error is thrown because the panel fails to launch.  The reference to the pin is in my postgui_hal file.

How do I diagnose and repair this issue.

Regards,

Ok, graphical.target is correct. This tells me it is configured to start to graphical, but apparently something is breaking it.

Lets try:
(run this before startx, then let me know what it answered:)
sudo lspci -nnk | grep -A 3 -i vga

maybe also run it again in a terminal after you ran startx and see if anything changed.

  I'm automating a parallel lathe with a 9-speed gearbox selected manually using two levers, each with 3 positions: lever 1 has positions A, B, and C, and lever 2 has positions D, E, and F. The possible combinations are as follows:   Direction changes are achieved by activating two electromagnetic clutches.This gives me the following truth table: I'm using the gearchange component for up to 32 speeds created by Andy Pugh, which can be used in this case. The first problem I encountered was creating the logic to detect these combinations, as I needed to detect the positions of the 6 switches I'm going to install to control the lever positions. Of the 6 bits, two will always be true. The component I found and was able to use was LUT5, but the problem is that it's only valid for 5 bits, and there are 3 combinations where bit 5 is one. Therefore, for these 3 combinations, in addition to LUT5, I had to use an additional AND2 gate to detect them.I read on the forum that for some automatic gear shifters, like the Maho MH400E, they created a component with an array that detects each gear by the switch positions.I also used the generic mux component to identify which gear was selected using a number between 1 and 9. Since it's 9 bits, it gave me a combination of 512, of which I only used 9 so that the mux output was a number between 1 and 9.I don't know if there's another way to detect the switch combination for the two levers.I've attached my HAL file to detect each lever position so I can send the correct speed to the VFD.

Hi Bari,

I came across your February discussion about Xenomai 4 and LinuxCNC and just wanted to provide a quick update, especially regarding Mesa.

In the meantime, Hannes has implemented Xenomai 4 / EVL support in LinuxCNC, as well as OOB networking for Mesa HostMot2. PR #4199 was merged into master on August 5, 2026.

More information here:

forum.linuxcnc.org/38-general-linuxcnc-q...core?start=10#348629

So the question of Mesa support that you mentioned back then seems to have been resolved now.

One option is to run LinuxCNC in a virtual machine on your Windows PC. You can use the simulator (SIM) configuration, so you don't need any actual CNC hardware. This gives you the same G-code preview as in LinuxCNC, including things like G83 peck drilling and the individual pecking steps.

Basically just run LinuxCNC in a VM in simulation mode.

Hi cgy,I came across your old discussion about Xenomai 4 / EVL and LinuxCNC and thought you might be interested in how things have progressed since then.Using Hannes'repository, I managed to get the complete Xenomai 4/EVL + LinuxCNC stack running: github.com/hdiethelm/xenomai4-linuxcnc LinuxCNC actually reports:So this is not just an experiment or a build that happens to work — I have it running on my actual machine.Hannes is getting really impressive results, especially with an Intel I350. I bought an I350 recently and I'm still tuning it, so my results are not nearly as good yet. But things are moving in the right direction.My current results and histograms are here: github.com/LinuxCNC/linuxcnc/pull/4087 Another important development is Hannes' PR #4199: github.com/LinuxCNC/linuxcnc/pull/4199 It addresses Xenomai 4 EVL OOB networking for Mesa HostMot2 and was merged into LinuxCNC master on August 5, 2026.So, just a quick update after all these years: Xenomai 4 + EVL + LinuxCNC really does work now, and it can be used with a Mesa card. ?Hannes has done a tremendous amount of work preparing the Xenomai 4 environment and integrating the necessary support into LinuxCNC.

No worries Greg. I have not used Qelectrotech before but for my wiring diagrams I use old school 2d AutoCAD. I looked at Qelectrotech and it looks good. I have used TinkerCAD and CircuitLab for some university courses I have been studying and they work ok, but I am sure you could do some good drawings with Qelectrotech.

Right, and up there is also the value of 7.3728 inches. But one thing does stand out to me... the distance to the tabletop remains virtually constant when I move along the X and Y axes.

This was without j2 and j3 +- 5.49
J2 upper arm J3 lower arm.

J5 5. Deg offset:

youtube.com/shorts/0iBH6llVtzk?is=Uy8PDOtG_uMjC4hk