Advanced Search

I hade the same issue with GladeVCP an Jogging already years ago. I was not able to fix it permanent: forum.linuxcnc.org/21-axis/33925-glade-p...is-blocks-arrow-keys

It made me finally give up with GladeVCP and "downgrade" to a good old PYVCP sidepanel.

Best wishes

Jan

I wrote a lengthy bit on tool offsets here:
forum.linuxcnc.org/38-general-linuxcnc-q...ets-be-stored#342300
it might help a little. Took me a while to "get it"
What I ended up doing is probing each tool and subtracting an arbitrary absolute Z height from the absolute probe Z, so every tool has a "length" that's relative to an arbitrary spindle position. It means no more reference tool, but it also means I always measure tools when I pick them.

OK so I do need help ive tried using this gcode to setup tool measurement but not getting accurate results from first tool I zero off work coordinates go through the gcode but the next tool comes out a few mm above work piece maybe more i didnt mesure the distance but in the gui window its at zero and in the G offset its zero.

What am I dong wrong ?  Please can someone give me the correct way of doing this.

Thanks Fred



g49

(measure reference tool)
G91
g38.2 z-190 f250 (measure)
#1000=#5063 (save reference tool length)
(print,reference length is #1000)
G90
G0 G53 Z0 (Retract to safe height)



(msg,insert another tool)
M61 Q

(measure next tool)
G91
g49
g38.2 z-160 f250 (measure)
G10 L1 P#5400 Z[#5063-#1000] (set new tool offset)
(print,new length is #5063)
G90
G0 G53 Z0 (Retract to safe height)


 

hallo an alle da drausen!
ich bin neu hier
kurze erklärung bin von mach3 auf linux cnc umgeschtiegen, hab mir ein plasma cnc schneider gebaut für hoppy anwendung dieser machte mit mach3 über usb port immer wieder probleme beim starten, (hochfreqenz e klar), daher jetzt linux über pararell port mit einem "TB6560-3V2 china port
soweit so gut funktioniert alles wunderbar, nur das relais aus pin 14 laut port funktioniert nicht habe schon alle versionen probiert , endweder es zieht immer an oder halt nie...
hat jemand eine idee woran das liegt?
danke im voraus

Just so we are on the same page:

1. Are you sure your tool directions when jogging are according to the right hand rule ( tool movement as seen relative to the table)
 

2. After homing the machine which end of the axes is the tool located (ie is it at the positive/negative end of the machine axis)

Ok, I found my one mistake. i was trying to compile "5axiskin.c" but it didnt work so after reading more, I found there is "userkins.comp" file need to be compiled. so i edited file as per attachment and getting error as per screenshot.


 

Hey everyone,
Has anyone been successful in configuring an XYYZ setup and would be happy to share their .ini and .hal files, and possibly the EC500 JSON config file (if modified)?

I've reviewed the suggested example files and mine seem to be correct for a dual Y axis motor gantry, but it's not working correctly and I'm chasing my tail trying to figure it out, so time to ask for some help! If possible!

Are you still stuck? I have some python scripts to load the firmware onto the board using Windows and powershell. Sometimes you need to power cycle the board a few times to validate that it connects.

Pardon my brevity,

Here is detail

I'm trying to compile this kinematics file "5axiskinsN.c" and getting this error

cnc@cnc:~/linuxcnc/configs$ sudo halcompile --install 5axiskinsN.c
Compiling realtime 5axiskinsN.c
Linking 5axiskinsN.so
objdump: section '.rtapi_export' mentioned in a -j option, but not found in any input file
/usr/bin/ld:5axiskinsN.ver:2: syntax error in VERSION script
collect2: error: ld returned 1 exit status
make: *** [/usr/share/linuxcnc/Makefile.modinc:124: 5axiskinsN.so] Error 1



 

In the screenshot of the first post the view is that of the real machine, with x0 y0 at the top right but y moving away becomes negative and if I modify the axis to make it positive, home moves to the left and the backplot is mirrored.

You are looking at the wrong Z.

None of those on the pictures are his.

 I think ArcOK is there.
My reading is 1 and 2 trigger the torch, 3 & 4 should  be ArcOK

Hello,

We created the `toolchanger_index_new` HAL file to map all the M-codes for `toolchange.ngc`; this kept things reasonably organized and allowed us to track what codes like M66 actually do.
Then we built the NGC file.
Once that was done, we integrated everything and added the remap section to the INI file.
When we started LinuxCNC, we got the error message mentioned in the previous post.
We suspect there might be an issue with the file path or something similar.
As soon as we comment out the remap configuration, everything runs without problems again.

The only exception is the `conv-float-32` converter; it used to work, but now LinuxCNC claims it already exists.
That is likely a separate issue and unrelated to the mapping setup mentioned above.

Tom, I tried everything, inverted the home encoder phases and all the possible settings... if I set the correct axis (reversed) everything is fine but the plot is mirrored and home moved to the left outside the frame.