Advanced Search

  Hi everyone! Follow-up to my earlier post — kept working on the profiler with Claude (AI) and testing on my real lathe (LinuxCNC 2.9.8, Mesa 7I96S). Found more real issues since last time, including a couple more crash risks, plus added several new features.What was fixed:

• G72's D/I/R were also wrongly scaled for G7 — same class of bug as the D/I swap from before, but G72 works along Z, where no diameter/radius conversion should apply at all
• Retract (R) in G71 was missing the same G7 scaling fix that D/I already got
• diagonal rapid moves could happen between multiple finishing/spring passes — now explicit, safe L-shaped repositioning between every G70 call
• a profile point sitting exactly on the centerline (X=0) could hang LinuxCNC before any motion even starts — now caught with a clear error instead

New features:

• tool nose radius compensation (G41/G42) support, including multiple finishing passes under compensation
• separate "spring passes" (repeat the final cut N times with no material removal) as its own stage
• finish allowance entered per pass (e.g. 0.25mm × 2 passes), with the total shown live
• ΔZ (enter coordinates relative to the previous point) and ⌀X (enter the desired diameter at an arc's peak instead of the radius)
• Half/Mirror preview mode
• full project import/export (.txt with all machine parameters, not just coordinates), with auto-detection on import
• Coolant/Tool comp/Cycle as dropdowns instead of text/checkboxes
• generate-then-save workflow (Save/Send stay disabled until you click Generate)
• better fill rendering in preview (incl. ID bores, and stock length the profile doesn't reach)

File and detailed changelog (CHANGES.txt) attached. As always — I'm not a programmer, so please double-check everything on your own machine before trusting it blindly.

I can't help you.
I gave you working example.
You only have to remove from your config unnecessary stuff and replace parport pin by mesa pin.
But you see only that is one line not many...

saxa wrote:

I was wondering if I should buy direclty the 7i95T as it has all needed i/o and encoder feedbacks for up to 6 axis.
What you suggest ?

Yes, that would be better, but finding it sometimes not easy,
With Mesa, you can add inputs and outputs and PWM and encoder feedback as you see fit:
-more IO = 7i84 or 7i70 or 7i71 or 7i72
-more encoders = 7i85 or 7i85S (not the same, only half the same)
-more PWm for +-10V servos = 7i83

For such a large machine upgrading from the 7i96s would be a smart move as io is limited (but you can add extra boards). eg. Depending on your needs 7i76eu, 7i95t, 7i97t.

Note that if you do fall back to ethercat, with beckhoff, you just need to add an extra input or output module at any time as required at a cost of about $50 for another 8 ports.

Don't forget LinuxCNC can handle about 8 spindles so the drill could get a free ride on another Z axis stage if you are creative.

Does your table have downdraft zones?

Yes, there will be synchronized movement as its all controlled by Linuxcnc on the single servo thread

halui.mode.teleop bit inpin for requesting coordinated jog mode 

tommylight wrote:

That would be redundant as LinuxCNC is the motion controller.
In short:
-buy a Mesa 7i96S and a THCAD

I was wondering if I should buy direclty the 7i95T as it has all needed i/o and encoder feedbacks for up to 6 axis.
What you suggest ?

I plan to set up a separate head in the future for plasma and who knows maybe even a spindle for drilling.

looking forward to your TP2. I would consider a few jerks in feed override an acceptable loss for using override as long as the max degredation of velocity is trapezoidal. Even a slight violation of max acceleration of an axis during the feed override is an acceptable degradation in my opinion.

If you are looking for a discussion on switchkins, I can help. When I first wrote switchkins, I could not think of an easy method to force a interpreter sync and yield type of queue buster which stops readahead. So looking at Fanuc and Hass, I added G12.1 and carried it through the complete chain of interpreter and motion. However dgarret came up with a much easier way to affect the queuebuster using M66 and remap which avoided touching so many files in the source tree for the switchkins feature.

For using axis limits during planning, the tool length which will become a parameter for the kinematics (including kinematics on the user side), could be carried through HAL or as a status variable from the #variables (#5401 to #5409).
I am looking forward to your release.
-automata

I wonder if there are two interfaces. For example, the X and Y axes are controlled by the Ethercat and the Z axis is controlled by a Mesa card, will there be synchronized movements?

scsmith1451 wrote:

I uninstalled Glade and then reinstalled it. The issue still persists.

I can't help you, but have a suggestion.
The "object has no attribute" is python's error and means "python can not finds", probably due bad path.
For me often helps to install python package in both ways, deb package and pip package.

forum.linuxcnc.org/9-installing-linuxcnc...e-official-amd64-iso

Yes. When I try to upgrade,I one part of the error message reads:  raspi-firmware: missing /boot/firmware, did you forget to mount it?.
I cannot post full message because the system clock is wrong so internet is not working.........

I'm using the physical keyboard keys to control jogging.

Checking halui.mode.is-manual = FALSE. I attempted to set it to TRUE but halshow reported that the parameter is not writable.

Is there a way to force this pin to be high?

Regards,

Sometimes a custom component simplifies your hal file. Here is my solution for a similar problem where there are two belts each with 4 positions. But because it's not possible for both belts to use the same position, it comes down to the same problem you have. For the ratios, I just ran the spindle on each belt combination and read the actual rpm with a cheap laser tacho and did some maths. But I did not have levers, and made a panel for qtdragon which Chris "borrowed" for the qtdragon docs.

github.com/rodw-au/vmnmillqt/blob/main/scalebelts.comp

Sean Drummond did a really nice diagram of a 7i96s which you can grab over at plasmaspider plasmaspider.com/viewtopic.php?p=231679#p231679 

If you want a diagram you can't go better. I usually just create a table of all the pins and where that go. Its better to do it in a word document rather than a spreadsheet so you can add more freeform text/