Advanced Search

bladekel wrote:

I am using a Raspberry Pi 5 with the LinuxCNC Debian Bookworm image.
I have configured the system to communicate over SPI.
I can move all axes, but when I run a G-code file, I experience a drift in all axes.

To troubleshoot the issue, I wrote a simple G-code program that only moves the X-axis: it moves 10 cm in the positive direction and then 10 cm in the negative direction.
I set this program to run continuously in a loop. In every cycle, a small shift occurs in the positive direction.
After 20 repetitions, I observed that this drift reached approximately 5 mm.
The start and end points of the program shift slightly in the positive direction with each iteration.
I have also tested it at very low speeds, but the result remains the same.

I am certain there is no mechanical backlash or electrical interference/noise.

I am attaching my configuration files. What do you think could be causing this?
 

For those who might encounter this problem, just like me;

This problem is caused by the dir_delay. 
Solved by increasing the dir_delay value.

Hi Zbigi,

Of course, happy to help! Please open an issue on GitHub:

github.com/cnc-proton/nativecam-py3-gtk3/issues

Most likely you hit the critical bug fixed in latest commit —
duplicate option in cfg/lathe/material.cfg caused a crash on
Python 3.13 before the menu bar even appeared.

Quick fix:

sudo sed -i '/\[PARAM_Z_RAP\]/,/^\[/{/^header = h3/d}' \
/usr/share/linuxcnc/aux_gladevcp/NativeCAM/cfg/lathe/material.cfg

sudo cp /usr/share/linuxcnc/aux_gladevcp/NativeCAM/defaults/lathe/default_template.xml \
/usr/share/linuxcnc/aux_gladevcp/NativeCAM/catalogs/lathe/projects/

Then restart LinuxCNC. Let me know if it helps!

   
Update: DRO is now showing only XYZB.

Edited dros_user.ui and offset_dros_user.ui removing the A and C rows and fixed the ZERO ALL command.

Added to INI:
DRO_DISPLAY = USER
USER_DROS_PATH = user_dro_display/
OFFSET_COLUMNS = XYZB

Screenshot attached.

The A+/A- jog buttons are still present — is there a way to remove them without touching system files?

Regarding the wrapped rotary test branch — ready to test. My setup is a standard XYZ milling machine with a B axis mounted parallel to Y. Where can I get the test branch and how do I get started?

I am using a Raspberry Pi 5 with the LinuxCNC Debian Bookworm image.
I have configured the system to communicate over SPI.
I can move all axes, but when I run a G-code file, I experience a drift in all axes.

To troubleshoot the issue, I wrote a simple G-code program that only moves the X-axis: it moves 10 cm in the positive direction and then 10 cm in the negative direction.
I set this program to run continuously in a loop. In every cycle, a small shift occurs in the positive direction.
After 20 repetitions, I observed that this drift reached approximately 5 mm.
The start and end points of the program shift slightly in the positive direction with each iteration.
I have also tested it at very low speeds, but the result remains the same.

I am certain there is no mechanical backlash or electrical interference/noise.

I am attaching my configuration files. What do you think could be causing this?

Regarding DRO — I already did exactly that, edited dros_user.ui removing the A and C rows. DRO now shows only XYZB. However the A+/A- jog buttons remain — they are controlled separately from the DRO file. Is there a way to remove the unused jog buttons without touching system files?

Regarding wrapped rotary — very interesting! I am ready to test. What do I need to get started with the test branch and what machine model is required?

You will probably find a 7i96s is cheaper in the end but there is a lot more wiring to do. Ethercat encoders are expensive and probably difficult to configure if you can't close things on the drive. Either way, try to get an Intel NIC. The balance probably falls in favour of Mesa if you can get away with a 7i96s

Except for versa probe and basic probe, the short answer is no general way to customize widgets.
Versa probe and basic probe look for customized ui and handler files.
This describes the process for them:
linuxcnc.org/docs/devel/html/gui/qtdrago..._probe_screen_widget

You can always add your own widget to qtdesigner.
the plugin link should be in a hidden file in your home folder:
.designer/plugins/python/Link to qtvcp_plugin.py
you could add more links, but it's a fair amount of work.

What widgets do you wish to customize?

Bumped into this, seems nice for understanding what does what in PID tuning, link is in description, web based so no install needed.

Yes that sound more reasonable thanks for that info I wanted to go by ethercad to have a small HMI PC for Linux CNC but if mesa card makes more sense sure than I go that way.

Can you maybe give me some info what to look for I need a computer with PCi port and the the card is the driver or do I need extra drivers and than the stepper motors ?

Maybe someone can recommend me a setup that would be nice thanks so far for the help I really appreciated it 

The commanded C position should come directly from LinuxCNC, It should not have steps
nor be affected in any way by PID settings.

Can you post your HAL/INI files?

 

do you have the plot turned on for machine bounds in the ui? sidebar tab "PLOT"

for the dro display, you can use the user custom dro template and setup however you desire, just remove the rows for a and c and leave xyzb

for the wrapped rotary, this will be something coming down the line, we are currently working on multi axis plotting and simulation so if you have a model of your machine and would like to be a tester for the currently existing test branch you can give it a try if you like, the feedback would be good for us to have someone outputting multiaxis gcode for testing.

Hi Axemas
Thank you for your hard work. As you suggested, I’ll send a request via GitHub. I’m having trouble with the lathe tab—can I count on a reply and help? Thanks, Z

Good day everyone,
My name is Radko. I got a secondhand MAHO MH 800 C too and having some difficulties with the mill.
I am need of the electrical schematic diagram of this machine.
If anyone has a soft-copy (pdf, image file) of the schematics then please send me.
If there is a link where to get the scheme
I would be very grateful if you could.
Best regards,
Radko

Thanks; am really looking for someone I can pay to configure this system, then help with programing of the router to make a number of parts.

In Zeile N55 fährt das eine file auf X11.57 und das andere auf X-1.444, die Differenz zwischen den beiden Positionen ist viel mehr als der Radiusunterschied zwischen D3.175 und D2. Da stimm evtl was in deinem CAM oder mit den Werkzeug Durchmessern nicht.

Gruse
Max