Advanced Search

I searched through the documentation before starting this thread. That only shows how to wire it to a raspberry pi.

As I said at the beginning I do not want to use a raspberry pi. I want to use a w5500 ethernet board. Then connect with a pc. Like shown in the demo video.

Futterama
Thanks for your error reports and certainly for the corrections.
I >think< Tommylight usually amended the link in Post #1 to the latest version. Unfortunately I am still not afloat. There is a 95% finished update of the script with quite a few amendmends. Not good to go public yet. I do hope I could give this some time soon.
Meanwhile, please do as you please and keep posting your findings.
greez
chris
... and thanks for the flowers anyway

I have a similar setup on my lathe; I'll send it to you later.

forum.linuxcnc.org/27-driver-boards/4698...7i96s?start=0#252760

I made a lathe with manual control of two axes simulatively, although apart from the servo motor as the main spindle I didn't have another axis as in your Z that could do its thing while these two are moved manually as I understood. maybe something in this topic can help.

this won't work either in sim mode or in reality, I tried something via chatgpt but I know that it doesn't work, it can't create a routine. here is my configuration if anyone is free to look at where the error is and what needs to be added in Hal eventually. Tx m6 works but will not return the tool to its pocket when you give it another tool, it also reports in the bottom line that there is no tool even though in reality it took it. each pocket has its own active high sensor on inputs 16-21 and the spindle has a sensor for ``flush'' input 10 the tool is in the spindle, input 12 reports that there is no tool in the spindle. Thanks 


youtube.com/shorts/CXxlm92xtt8

Sorry for my bad English 
Here are a picture from the handwheel /mpg and from the mess/cabinet

Not sure if it will help a lot 
Pg up and Pg down buttons works well however I need to manually select the axis. The knob doesn't change the selected axis in Axis gui

Edit: I finally found the issue, I always struggle between axis and joint... 
I managed to move axis using the joint instead of axis. 

I'll try an make new firmware sometime today or tomorrow.

What I will do is add combined encoder/stepgen index pin to the first 4 stepgens
on 7I89 encoders 0..3

There are some constraints with muxed encoder/stepgen index pins so
5 is awkward but  6 is possible if really needed (with a dummy stepgen)

5i25

Ok, thanks, I'll try the G540.

It that a 5I25 or 5I25T?
 

adding:
        self.w.jogincrements_linear.setEnabled(0)

to the end of
    def init_widgets(self):

Will make it not respond to screen clicks, but should still respond from HAL pin changes

I have solved this. I had a similar issue. Trick is to upgrade to 2.10 by compiling your own version. I highly recommend using ChatGPT or Gemini to help with this process. This has a new feature for pipe cutting.

What you then do is to enable pipe-cut in the HAL-config OR enable it through Hal Show, this avoids the Z axis movement probe cycle prior to running the program. I can now run my XXY with a fixed Z and get the torch to run without needing to do the probe cycle.

I also created a button in the UI under settings to toggle the hal pin pipe cut to make this easier because the flag was reset each time i loaded a new program.

I've taken a look and haven't found such an option for this particular config. 5i25+7i76u+7i89. Would that need custom firmware? I see a firmware with index for 7i95t, so maybe I should've gone that route from the start.

I have a XXY plasma without Z axis. Similar issue. I compiled a new version of LinuxCNC 2.10 then I set the tube-cutting in HAL to true in my .HAL. This avoids doing the Z probing before starting the program.

I did notice that once once I load a new program, it resets the tube cutting to False to I created a button in the QTPlasmaCNC UI under settings which I named IGNORE Z-PROBE and then you can look at the config how to toggle a pin to set the tube-cutting to True.

This seems to work well in converting a non-Z plasma cutter into a working state but still use QTPlasmaC.

You might post a link to this thread on the MesaCT topic