Advanced Search

Aktualizacja sytuacji: Okno GUI AXIS jest odcięte ZARÓWNO na dole, jak i po prawej stronie.

Rozdzielczość mojego ekranu to 1366x768 (LinuxCNC 2.9.8 na Debianie). Ponieważ 1366px powinno być wystarczająco szerokie dla standardowego AXIS, wydaje się, że zintegrowany panel boczny (jak PyVCP czy GladeVCP) wypycha główne okno poza widoczny obszar ekranu.

Czy mógłbyś mi powiedzieć, jak poprawnie przeskalować okno lub zmienić pozycję panelu, żeby wszystko zmieściło się na monitorze 1366x768? Czy powinienem szukać linii PYVCP lub GLADEVCP w moim pliku .ini?

I now have X and Y working on the actual machine and I want to give this another go. Can you clarify if to use cia402 homing I need to have the limit switches wired into the servodrive specific input pins, or if it works all the same with them wired to beckoff input pins?
Yesterday I had it sort of working (motor direction wrong + homing direction wrong yielded something sort of working) with them wired into beckoff but i'm wondering if by any chance it could be using regular homing because of something else in the config.

Mesa 7i97T would be nice if you can find it, or 7i92T with 7i77, or other Mesa options.
In the mean time, have a look at some of the retrofits i have done with such boards, and what it takes to endeavor in such undertaking:
forum.linuxcnc.org/30-cnc-machines/33529-hurco-bmc-20p-retrofit

forum.linuxcnc.org/30-cnc-machines/31792...-01-retrofit?start=0

And when you get to wiring and tuning, this is a must:
forum.linuxcnc.org/10-advanced-configura...ning-detailed-how-to

Anything and everything, the original post does not translate to anything useful.

Hi everyone! I'm not a programmer, I have a small lathe (LinuxCNC 2.9.8, Mesa 7I96S), and I'm slowly learning. I found a few real issues in the profiler — some of them can actually cause the tool to crash into the workpiece. All fixes were made together with Claude (AI), and tested on real hardware.What was fixed:

• English/Ukrainian language switcher at the very top of the window (English by default)
• D and I were swapped in G71/G72, plus incorrect scaling for G7 (diameter mode) — the actual depth of cut was twice the value entered
• unsafe G71→G70 transition — for certain Start X values the tool wouldn't retract and crashed into the part
• Start X / retract didn't account for arcs (G2/G3) "bulging" beyond their endpoints
• added a "Finish only" mode — re-run just the last pass without G71/G70
• added a safe approach move at program start, from any starting position
• improved preview (symmetric, centerline at X=0) and DXF import
  
  File and detailed changelog (CHANGES.txt) attached. If it's useful to anyone, feel free to grab it.
   
  

I have been looking to implement something like this on my machines. I downloaded the ngc files that you have posted and read your write up on it. Would it be possible for you to post the python files. I believe this would be a great option to be added to the main repo for makers like me. There are a lot of machines out there that have this feature in the controls already. Examples are Onfinity, Sienci Labs and others. It fills a huge void in semi automatic tool change area. I switch my machines to Linuxcnc do to replacement cost after controller failure and the options available with linuxcnc that other controllers don't have. Thanks for all your great work! 

interactive version:
forum.linuxcnc.org/41-guis/58749-halview

If you don't want to replace the servos, go with Mesa.

Hi,

I'm considering retrofitting old (90's) milling machine, industrial type. For now i know that this machine uses -10V..10V to drive spindle (with internal tacho) and -10V..10V to move axis (XYZ). It has magnetic linear encoders for each axis. Given that i have a few questions:
- should i use mesa cards or bechoff ethercat?
- what kind of PLC logic I will need to write (lube, coolant, safety) afterwards?
Any kind of help will be appreciated.

Thanks everyone! Once I install the Sanmotion AC servos, I'll see how to configure them properly.

Sorry it took such a long time. I have updated the code with your suggestions, pushed it to github, but have not gotten around to testing it yet on my own lathe as workshop time is scarce at the moment.

Let me know if you still encounter any issues.

1. As you have already found out 'halcompile' will not work with all .c files. A 'run-in-place' (RIP) installation would be required here.
2. .comp files have their own syntax, so you cannot simply copy-paste the code from a .c file into a .comp file.

Linuxcnc will let you do anything but there is an expected paradigm. If you go against the paradigm, life becomes painful.
Think of the home position as being 0,0 on those Cartesian graphs you drew at school and the table represents one of those quadrants.
So if the current position is home, X & Y need to be  negative values to move away from home.

Another way of looking at the expected paradigm is that the front of the machine is where you stand at the machine. Home should be on your left and closest to you with the gantry going across your chest. So to satisfy that rule, you need to stand at the back of the machine behind the gantry.

So you could
1. stand behind the machine
2, Turn the machine around so the back as shown is the front of the machine
3. Live with negative X & Y coordinates
4. Move your home switches to the front left of the machine
5. In your homing configuration, set the home position to max-X and max-Y coordinates so it knows the home switches are at the back of the machine.

I might add if your origin/home respects the cartesian origin and you accept the quadrant you are in, it all becomes moot when you place your workpiece on the table and set the part origin with G54 offsets. Once you touch off, everything is referenced to the part origin.
 

For basic information on your configuration you need to post your .hal and .ini files without that we are just poking around in the dark.

Es gibt zig Dinge die schief gehen können.
Besonders mit (billiger) Elektronik.
Das Durchmessen vom LPT Kabel ist da wohl eher der Anfang, als das Ende.

Wenn die Motoren funktionieren zeigt das ja erstmal nur, dass es grundsätzlich geht. Und Linux startet.
Was war denn jetzt mit der Möglichkeit, dass das Relais "klebt"?

Ich würde mich jetzt von einer Seite durchwühlen.
Entweder aus Richtung Relais, oder aus Richung Computer.

Was ist denn schon genau ausprobiert. Ich würde mir eine Liste machen, was ich alles schon probiert habe. Die kannst du dann auch hier veröffentlichen dann geht es evtl. weg vom bloßen Raten.

a) M3 mit Drehzahl (siehe Tommylight)
b) hal show, nachsehen, ob Linux CNC den Pin am Parallelport versucht zu schalten
c) Led mit Vorwiderstand oder mit Spannungsmesser direkt am LPT messen, ob das Signal hin- und herschaltet
d) LPT Kabel wieder anschlißen, selbe Messung wiederholen
e) Computer abstöpseln und alle Verbraucher von der Karte abstölpseln und probieren, ob das Relais mit Spannung am Steuerpin schaltet.

Danach müsste man doch wissen, ob es auf der PC Seite hakt oder mit der Steuerkarte was nicht stimmt.

Gruß Timo