Advanced Search

Hi everyone,I'm exploring the possibility of retrofitting an older Reishauer RZ300E gear grinding machine by replacing its existing electronic gearbox with a LinuxCNC-based solution. I’d appreciate your thoughts on whether this is technically feasible—especially for someone with limited experience using LinuxCNC.Additionally, I came across this article on continuous generation grinding  and video() that outlines some of the machine’s principles. Could a setup like this be simulated and validated in LinuxCNC before committing to hardware implementation?Any guidance, suggestions, or shared experiences would be greatly appreciated!Thanks in advance.

I think it's hardcoded into stepconf, otherwise it would have to parse through known editors, there are many and someone would complain (vim & emacs I'm looking at you.)

Alot like geany so that was the editor chosen (same for the arm64 images).

Python3 has been the norm for a while now. I think the gcode generators are 8 years or so old.

There is a lot of python 2 stuff around that just hasn't been update for python 3.

You could install python2 and try that for the gcode generator.

Yes, you are correct. The way the python script creates the gcode in this particular script is by using trigonometry to find the coordinates and writing to sys.stdout. I did find a preamble and postamble in the py file and I found the postamble was correct with an M2 so I think it has something to do with the way it sends the code to Axis.

Thanks Tommy. I had tried G93 Inverse Time, but didn't see any difference to G94 Programmed Feed.

I went back and did a bit more testing (with G93 and G94). The only thing I've found to work (for both G93 and G94) is to up the velocity and acceleration speeds to unachievable levels, or lengthen the small line segments that make up the toolpath. However, this will result in a more inaccurate part. And I still can't get close to the required feedrate while keeping the short lines to a decent length.

Thanks for the heads up! I don't think that would affect whether the mdi-command-NN pins would show up or not; even if don't load the PyVCP panel, I still don't see the pins exposed by halui.

Fresh install and then run stepconfig and noticed that if I try to edit a loaded gcode file in Axis, it does nothing. So I take a look at my .ini file and it points to Gedit but that isn't in Debian 12. It uses Geany. I installed Gedit since it is what I'm used to. I could change it to Geany but I like Gedit better for gcode. 
 I just thought it strange that stepconfig would use a non-existent editor.
 Python 3 also makes the simple gcode generators not work anymore and I haven't figured out how to get Native Cam to install.

See if this helps, although pretty old.

"No percent sign or M2"
Means there is no % at the start and end of the gcode file, or M2 at the end.
See if those can be added as a header and/Or footer, or maybe as post processor?
All this is a stab in the dark from what i know using Inkscape where both above options work fine for LinuxCNC.

I had to use "2to3" to try and convert them to Python3 and was able to get the GUI's to work on Bezel and a few others and then I filled out the the dimensions and hit "Write to Axis" but it always errored with "No percent sign or M2" and discarded the gcode file. I couldn't get them to write to the clipboard either. I couldn't get "Engrave" or "Bolt pattern" to run without "indentions or tab" errors.

Sorry, forgot to mention updated hal line as well to set the SPI speed to 2Mhz, the default 20Mhz is probably the issue.

loadrt remora-spi SPI_num=0 CS_num=0 SPI_freq=2000000

Gusgus wrote:

Hi again and thanks for all the help up until now!

I wan't to give more context and explain myself better, we have been using this cnc router since 2023, originally setup with a raspberry pi 3b and octopus pro combo, it has worked wonders and is amazing the only real issue is that the raspberry pi 3b was slow. Recently we made the change to a raspberry pi 5, we took the original .ini and .hal files which we then update to be compatible with the new remora version, we kept the same config.txt file. We installed the linuxcnc-raspberrypi5 image and installed the remora components as instructed in the documentation.

The issues we've seen: when we first loaded linuxcnc and ran the .ini file by double clicking on it and choosing linuxcnc as the app to run the ini file we then saw the raspberry pi 5 using rp1 driver alert, we initially thought this was error but figured this was just an alert and so continued. We then started testing leaving e stop we could but just for a moment, no more than 3s, we could even get to enable but non of the drivers actually enable. We then tested if this was the E-stop's fault but we assumed it wasn't since if we disconnected the estop we could exit estop at all, wehn it was properly setup we could but can't enable the steppers fisically. This isn't a stepper issue or driver issue since they get enabled with the raspberry pi 3.

Thanks in advance and thank you for all of the help until now! Oh and sorry for any grammar or spelling mistakes, english isn't our first tounge.

Ok, so I pulled out my octopus board and RPi5 to see what could be the issue. I found my working config and noted:

#SERVO_PERIOD = 1000000
SERVO_PERIOD = 5000000

Making this change to your config, I got it up and running. With the faster servo period the board was throwing so many communication errors.

There is  </vbox> missing at the end of of the .xml file

Sorry to bother you again guys. With the help of some people I have the axes and the spindle going now.
My current issue is the homing of the axes. They do not move to search for the home- switches, but just home "in place" so just where they stand. Obviously the axes pretend to "be homed" then, because softlimits are active. Signal joint.1.homed swaps immediately to yellow
If I unhome all axes I can move them where I want by jogging them- otherwise the + way is limited as it should be to 5/10mm +.
The hm2_7i80.0.gpio.026.in reacts in HAL-show on pushing the home switch manually and also the signal  "z-home-sw" swaps from red to yellow together with hm2_7i80.0.gpio.026.in.
The signal "z-is-homed" stays red .... if this is of any importance.

It is just that the search for the HOME signal does not take place. When trying to home an axis the axis motor is activated, stops immediately (does not move the axis) and sets the DRO value to 0.0
Any help appreciated
Thank you
Sigi

Yeah, I commented those out to get LinuxCNC to start so I could look at the halui component in halshow, to see if maybe the pins were numbered differently than I expected or something. With those lines uncommented LinuxCNC won't start, since the mdi-command-NN pins don't exist.

Thanks for mentioning it though, I've certainly had forest-for-trees errors like that before where another pair of eyes spotted the problem immediately. 

Just in case you are not aware that in the # Syringe section in your postgui.hal everything regarding mdi-command is commented out.