Advanced Search

And the PC MUST be set to 10.10.10.nnn where nnn is anything from 1 to 254 but NOT 10.
So yes, the PC set at 10.10.10.11 will work.

Hi all,
I am a newbie to cnc and to linux so please be patient with me. I have seen some posts where using the 7i76eu board can be set to using ip address of 10.10.10.10 I see some guys say changing w2 and w3 manual says w13 and w14 default is 192.168.1.121 with both down. and both up 192.168.1.121+ use fallback configuration. Some guys say w3 and w3 off to set ip to 10.10.10.10. I am confused many say i should be using 10.10.10.10 and hve pc to 10.10.10.11. Does the 7i76eu magically change the ip with w13 and w14 in the default down and the w2 and w3 in down position, or do you have to do something else in progaming also.
This my seem stupid to some of you more experienced users but I am confused!
Thanks for any help

Hi! Great Hug from Brazil! A Questions? Who I'd Can Help me , I trying ramake PCB of the BCL-AMP Friendess Preamplifier, i need know techinical informations since Already thanks very much for atention

NOT my stuff!!!
Been seeing this for quite a while now, but this is a "how to use it" where several nice features can be seen that make it easier to use the machine.

My guess is Rufus is the problem, create a new ISO with Balena Etcher

From a quick glance, there are no offsets in the ini file set for joints/axis, there needs to be some small offset to move away from limit switches after the homing is done, usually 2-3mm, adjust accordingly for imperial units.
And wizard can not move both tandem axis at the same time, so using it to test tandem joints is not recommended till that feature is added.
Edit:
Also usually the offset value is negative, so -2 or -3mm

.hal/.ini attached:

1. yes, industrial ones are very expensive and very heavy, also fully equipped for any task. I would steer clear of china ones for under 20K as those are "projects", not tools.
2. Biesse, SMC, etc can be found for cheap and retrofitted, not easy and a lot of work, but definitely worth it as you end up with a machine built like a tank with controls that never fail.
3. There are some examples around here and a nice little piece of software for it, but i never used it so i'll leave that to others.

Check the hal file if there is anything using the base period, if not you can remove it.

Hi,
I tried Editor, Notepad++ and eq. all the same result!

Still trying to figure out what the problem is...

If I open the file with gedit from Linux, everything looks fine, but as soonas I change to LinuxCNC I have the same crap...

Freddie

I have a 7I77 and a 7I76 hooked to a 5i25. And then there is a 7I66-8 and a 7I66-24 connected via serial (but I don't think those matter).

I looked for a while and didn't find / couldn't figure out with firmware I would need. Can you recommend an exact firmware?

Thanks!

To get hardware PWM you need to have firmware on the FPGA that supports PWM.
For example if using a 7I76, there is FPGA firmware that replaces 1 Step/Dir output
with hardware PWM/Dir
 

Posting your hal/ ini files would help to pin down the issue.
 

Hi all,

I'm planning a gantry router build for machining composite panels into cabinet components. Looking for advice on machine choice and stack.

Material

- Sandwich panels: 1mm aluminium skins on both faces, polyurethane foam core
- Total thickness: ~60mm nominal, but varies ±1mm across a batch
- Panel size: up to 1500x3000mm

Parts and operations



- Cutting outlines
- Pockets and step joints on panel face
- V-scoring the aluminium skin to create precise fold lines (score-and-fold)

These operations require multiple tools in a single job (end mills for pockets and joints, V-bit for scoring), so automatic tool change (ATC) is a hard requirement.

The step joints are the critical operation: they create mechanical overlap between mating panels so cabinets assemble with tight joints. A 1mm thickness variation in the stock translates directly into a gap.

Workflow

probing → generate Z map → process map (Python) → adjust G-code geometry (XY and Z) → send corrected G-code → machining

- Fusion 360 CAM produces nominal G-code
- Before cutting, I probe a grid of Z points across the panel surface
- A Python script reads the probe map and rewrites the G-code: step depths and joint profiles are adjusted to the actual measured local thickness — for both the panel being cut and its mating counterpart
- This is not just Z surface compensation: XY profiles are modified based on the probe map so that mating joints match regardless of thickness variation

Why LinuxCNC

- Workflow already tested with LinuxCNC + MESA cards (7i96/7i77) on an entry level router.
- Plan to use the Python socket interface for probe data collection and G-code rewriting — tight integration between probing and post-processing is central to the workflow

My questions

1. Are there manufacturers selling new gantry routers with these specs who ship with LinuxCNC, or sell mechanics separately from the controller?
   - Working envelope: ~1600x3200mm
   - HF spindle: 18k+ RPM
   - Integrated vacuum table with independently switchable zones
   - ATC (tool magazine or rack)
   - Probe input

2. For a retrofit path: which used industrial gantry platforms have the best LinuxCNC community experience?
   - Main concerns: ATC integration and vacuum zone management via MESA I/O

3. Any experience with probe-map-driven G-code modification beyond simple Z surface mapping?
   - Currently planning G38.x + Python
   - Is there a better pattern for this use case?

Thanks

 

Phew, its taken a few years, but my irrecoverable account has finally been deleted and I have been able to re-register under my usual username,hoorah!

While trying to troublshoot a CNC router that has been running just fine for years and suddenly can't find its 5i25 card ... it occurred to me ... does a Mesa based install even need a base thread?  Surely just a servo thread is all thats needed?

Can I comment out the BASE_THREAD in [EMCMOT], remove the thread start in HAL and move any HAL things that were on the base thread to the servo thread?

I w