Advanced Search

thanks, in that case I like the option to flash the card for 2xbob, both bob via a fmale to fmale 26pin cable?...in that case I need for my lathe machine one spindle analog output, one spindle encoder input and xzA config for servos and two independent simulative mpg wheels for x and z. In that case, will the pnccfg and gecko initial configuration give me the ability to meet the above needs, possibly by modifying hal for mpg and spindle encoder?

Yes, the G540X2 configuration supports step/dir on both connectors (8 stepgens total)
The 5ABOB x2 configuration is anothe possibility

You _can_ use the output directly with some limitations.

1. They are 3.3V outputs but can drive 5 V  loada when using in sinking mode
( say drive STEP+ to +5V, drive STEP- to  7I92 STEP)

2. The 7I92 pins can be damaged easily by overvoltages (input output range is -0.3V to +7V)
Its very easy to damage with negative voltages unless there is some current limiting

How do you get a negative voltage with high current? Just connect a GPIO pin to frame ground
that has 1VAC of of HF electrical noise relative to 7I92 ground from axis drives/VFDs
If you use the 7I92 GPIO directly, its best to have a small series resistor to protect against
this issue
 

need help with card configuration through pnccfg or mesact...namely specific question can 7i92tm be used without bob  cards directly and if so what is the initial configuration that should be selected in pnccfg? there is an option for gecko in the menu, is it adequate in case of using two bob cards or some other option. so that both P1 and P2 connectors are functional. thanks?

Wow, those are some pretty hefty files you're working with. 
Thanks for sharing more with us.

I have another question. I select the option to turn on the spindle. Output 27 is the option available, but it doesn't output 24V to turn on the relay and start the engine. The other available outputs (numbers 12, 13, 14, 29, 30, 31) always output 5V. Therefore, I wouldn't be able to connect a relay for the refrigeration pump and other control outputs. I've already changed the input and output functions, but the voltage remains the same.

At work I use a 5axis large format machine, ive had over 100mb files before cutting for ~25 hours. 
It has a tool changer but rarely used since it gets full of crud- crappy gantry design with very little dust extraction.
Much fun!

Hi everyone,I owe you an apology for the month of silence. The reason is that I've been completely focused on a hardware project called Proton – a CNC motion controller I'm developing alongside NativeCAM. I've put up a page atwith some info, but honestly, that project has been consuming all my focus and time, and I let this thread slip.Now that the hardware side is stable, I'm back and catching up on everything. I've read all three pages of this thread, and I've just replied to each of you individually (Surmetall, COFHAL, Doug Crews, and others).Over the next week, I'll push fixes to the GitHub repo (especially the blank GladeVCP tab and the OWord path documentation) and prepare the lathe test package.Thank you for your patience – NativeCAM is still very much alive.— Axemas

@Doug Crews – thanks for the report. The blank tab is a known issue with the -4 package on Trixie. GTK deprecation warnings break the embedded GladeVCP renderer.

Workaround: run nativecam standalone once, close it, then launch LinuxCNC – the tab should appear.

Also check your INI: NCAM_DIR should point to /usr/share/linuxcnc/aux_gladevcp/NativeCAM if the default doesn't work.

I'm fixing this properly in the next release (-5 or -6) within a few days.

— Axemas

Run one of the included configs for parallel port, if those run you are OK.

@COFHAL – you asked about the ~/nativecam directory and OWord macros. Here's what's going on:

NativeCAM creates and uses ~/nativecam as its working directory for temporary files, tool tables, and post-processor configs.

The SUBROUTINE_PATH in your LinuxCNC INI file must include the path to the OWord subroutines that NativeCAM generates. Without that, LinuxCNC won't find the macros.

If Axis shows no errors but QTDragon fails with "can't find macros or OWord," it's almost certainly a path issue.

Quick fix: add something like this to your INI file:

text
[RS274NGC]
SUBROUTINE_PATH = ~/nativecam:/usr/share/linuxcnc/ncfiles
Restart LinuxCNC and try again. If it still doesn't work, paste your full INI file here and I'll debug it with you.

— Axemas

@Surmetall – thank you for the kind words and for the continued excitement about the lathe features, especially the polyline workflow alongside G71/G72/G73. That preview video took a lot of work, and it means a lot that it sparked anticipation.

Since my last update, I've been testing the polyline path generation on real G-code scenarios. No major blockers – mostly fine-tuning the toolpath logic and cleaning up the UI response for lathe operations. I'm planning to push a test package soon.

And yes, I'd absolutely love for you to test it on a real machine. Your offer to provide feedback, screenshots, and G-code examples is gold. I'll reach out to you directly when I have something ready for real-world testing.

— Axemas

Hey @giunta.gio — apologies for the late reply, I completely missed your question back in May.The main hardware specs and overview are now live at cncproton.com/ We're currently finalizing the full technical documentation package. Over the next couple of weeks, the site will be updated with:

• Detailed wiring diagrams and complete pinout tables
• Example HAL configuration files for LinuxCNC
• More photos of the assembled modules and various configuration options

If you have any specific questions about your machine or particular hardware features, feel free to ask here or reach out through the contact form on the site — I'm happy to go into detail.Axemas

Hello Tommy:

The installation worked and I am able to launch the latency_histogram and 3D gears tests.

However do you know if there is a script or I am still missing steps in compiling the 
rtai_lxrt.ko, rtai_fifo.ko, rtai_sem.ko,rtai_sem.ko, rtai_tbx.ko, rtai_bits.ko and rtai_mq.ko kernel modules.

I check the /usr/realtime-5.4.258-rtai-amd64/modules and I only see a shorter list of compiled RTAI specific kernel modules
drwxr-xr-x 3 root root   4096 Jun 10 15:36 .
drwxr-xr-x 7 root root   4096 Jun 10 15:36 ..
drwxr-xr-x 2 root root  24576 Jun 10 15:37 linuxcnc
-rw-r--r-- 1 root root  26785 Nov 18  2022 Module.symvers
-rw-r--r-- 1 root root  38088 Nov 18  2022 rtai_hal.ko
-rw-r--r-- 1 root root  43368 Nov 18  2022 rtai_math.ko
-rw-r--r-- 1 root root  21368 Nov 18  2022 rtai_mbx.ko
-rw-r--r-- 1 root root  46816 Nov 18  2022 rtai_msg.ko
-rw-r--r-- 1 root root 268328 Nov 18  2022 rtai_sched.ko


thanks
Michael 

 

Always the 'joint 1 on limit switch error' or 'joint 0 on limit switch error'.
Mostly at pierce now, but has slowly been creeping in and would do it halfway through a cut. same error.

rodw wrote:

Spumco, great work. If you have to manually start the service each time. sounds like you overlooked the udev rules in the first post, Sorry about the formatting errors but its an awful post to edit

sudo geany /etc/udev/rules.d/99-ethercat.rules
and the following code block.
 

Rod - I hadn't actually gotten to that section of the installation yet.  The first 'ethercat slaves' check is before the 99-ethercat.rules line.

I just stopped when I got the errors starting ethercat service for the first time.  Bulldozing ahead past errors has not served me well in the past, despite my habitual attempts.