Advanced Search

It is an assumption. I click on "Write to Axis" and I notice that Axis says "filtering" at the bottom right but nothing is received. I then exit the py script gui after a few minutes of waiting for it to post code and that's when axis gives the message about the % sign or M2. It could be it never got anything.
I may change the "sys.stdout.write" to create a file instead and see if it post.
There may be something else in the py script that hangs it up. These scripts all ran fine on the python version that comes with Wheezy on my other machine but they don't like Python 3 and using 2to3 doesn't fix everything.

Not sure this will shed any more light, but I've attached the output from starting LinuxCNC on my actual machine.  I get that 'halui exited without becoming ready' message too.  One other thing I noticed is that the initial messages about kinematics has "coordinates:AZ" but later on the jog order is "ZA".  Is there some requirement that the first axis be linear?  I don't think it would hurt anything for my machine to reorder things so joint 0 was linear and joint 1 was angular.  Just kind of fumbling in the dark here for ideas.

 

No, no real help - but we found the error:
Wrong is the term VELOCITY - needs to be VEL thats all, then it works
so
Wrong:
HOME_SEARCH_VELOCITY = 10
HOME_LATCH_VELOCITY = 0.5
HOME_FINAL_VELOCITY = 1.0

Correct:
HOME_SEARCH_VEL = 10
HOME_LATCH_VEL = 0.5
HOME_FINAL_VEL = 1.0

Thats all, then it works

Sigi

That line does not show up if the axes are all linear.

JT

Dunno if this is a clue or not but <commandline>:0: halui exited without becoming ready seems odd.

JT

Hallo Forumsmitglieder,

ich möchte eine kleine Koordinatenbohr- und Fräsmaschine aus der Mitte der 80er Jahre mit LinuxCNC "retrofitten". Die Maschine hat 4 Achsen und verwendet "gebürstete" 12 V Gleichstrommotore von Excap..Die Positionsmessung erfolgt über 3 Heidenhain Glasmessstäbe und einen Heidenhain Drehsensor. Die vorhandene Heidenhain TNC113 bietet nur eine Streckensteuerung. 
Daher habe ich mir eine MESA 7i97T gekauft. Ein PC ist beschafft und Debian/LinuxCNC sind installiert. 4 Heidenhain EXE sind auch gekauft. Jetzt hänge ich an dem Problem, dass die MESA 7i97T nicht erkannt wird. Da soll man wohl die neueste Version von Mesaflash laden und installieren (nach Talla83). Leider sind die Videosequenzen der Programm-Eingaben etwas schnell und für mich nicht komplett nachzuvollziehen. Leider habe noch nie in meinem Leben irgendwo einen Programmcode eingegeben.

Vielen Dank an Talla83 für seine schönen Youtube Videos.

Ich suche daher Kontakt zu jemandem, der vor den selben Problemen steht und ein ähnliches Projekt angeht.

Ich wohne im westlichen Münsterland. Meine Email lautet: hartwig.droegemueller(et)outlook.de

Über eine Kontaktaufnahme würde ich mich sehr freuen.

Mit besten Grüßen

Hartwig

Yes, i think you are running into the single-line lookahead problem with rotary axes,

Tormach has 9-axis blending, but might not handle advanced kinematics as well (no joint/axis concept separation).

It would be nice to merge the Tormach 9-axis blend TP into Linuxcnc, but it doesn't look to be a trival task.

Which editor are you using?

Looking about on the internet it seems that maybe geany does odd thing in some locales. What's your locale setting?

(eg: unix.stackexchange.com/questions/340469/...ic-encoding-in-geany )

JT wrote:

Changing A to X and the mdi commands show up, changing back to an angular axis and the mdi commands don't show up.

 

Thats more than a bit wierd. 

Any clues from starting LinuxCNC at the command line? It sounds like it is giving up before parsing the MDI_COMMAND lines. 

Hello,I am trying unsuccessfully to create a parametric file for measuring the length and diameter of my tools.I am encountering significant issues with the encoding (UTF-8) of the subs. Either it cannot find the sub, or invisible characters are embedded in my files and LNC refuses to read them. Is there a solution, an editor capable of writing a clean file? Thank you.

Yes, you should change it. Lots of CAM software assumes that Z=0 is top of travel too. If the "safe move" instead rams the cutter into the table, that would be annoying.

is there any indication that the postamble is being sent?

Does it help if I say that I am as puzzled as you? Those config files look like they should do homing.

Is there any chance you are not running the config you think you are?

It might be worth halscoping the home switch signals, perhaps there is noise on those that looks like valid home switch-on home-switch-off ? (very unlikely, though)

johns00056 wrote:

Thanks for the help. The script to automatically re configure configuration files did not work..

Is there any chance you can zip-up the config folder and attach it here, I will try to work out what went wrong (and hopefully fix the script)

The Melfa seems to be all in-line so that makes the kinematics somewhat simpler.