Advanced Search

Hello

i am trying to setup ethercat on linuxcnc 2.9.4.
I use these instructions

when i want to start the service it gives an error:
 modprobe: ERROR: could not insert 'ec_master': Key was rejected by service

these are the conf params i set:

MASTER0_DEVICE="dc:4a:3e:52:df:de"
DEVICE_MODULES="generic"

i based it on this:
Does anyone know where i went wrong?
I did first install it and find out that i was on a sercure boot, so i disabled it in the bios.

Kind regards,
Andries
 

ok, thanks i check it out.
worked perfectly. 1,5mm difference. Is there a way to compensate for this directly in pncconf. the Y2 sensor is reached 1.5mm faster then Y, so Y2 should move 1.5mm further in the homing sequence?

I was wrong!. It can read rpm in both forward and reverse after watching spindle-vel-fb-rpm in halshow.

I was surprised at how close it was to the set voltage (based on the cut height setting.) Well thought out.

I have not played with this plasma in ages - it was initially setup by me before eoffsets.. It worked good - but this was much easier..

sam

Sounds like you may have the spindle ENA polarity backwards.
 

No, if you setup immediate homing, no home move or sensors are involved,
LinuxCNC simply takes the current position as home.

the problem is i want to be able to jog the gantry before the home all, so i can find out what the physical misalignment of my sensors is. if i home first the homing will force my gantry according the sensors and so misalign my gantry and see it as homed and square

After the last update the formating is fine now.

But the question i have now, about gcode filter. I tried the filter, i am using with axis, but nothing happened.
Is it possible to use a filter and my setup is wrong?

And another problem:
A subroutine for probing is called at program start. To prevent problems with tool not loaded yet and others, i test for #<_task> EQ 0 and return to mainprogram, if 0
But this makes the preview disapear, also after probing is done. until abort or reset the program. only then it comes back.
Is there an alternative probe button for testing?  to load  'simulate_probe &'  doesn't work anymore

edit:
after testing a dummy filterprogram, its definetely not possible.
About the preview, i commented all probing and stuff, just left the bare nc code. When starting the program, preview disapears, on reset it returns. Program is checked in the settings


 

Maybe setup  a test ini file with immediate homing so no
home move is done and then you should be able to jog the
gantry axis after you hit the home-all button.

linuxcnc.org/docs/html/config/ini-homing.html#_configuration
 

Thanks.

I soldered both resistors in the + und - Line inside the plasma source and also connected the 2 torch-start pins of the plasma-source with the Spindle enable + and Spindle enable - -Pin of the Mesa 7i76e-board.

When I power up the plasmacutter it instantly fires although the torch is not activated in plasmac. When I remove the 2 wires from spindle enable +- then the plasmacutter doesn't start when powered up.

When I measure between s-en+ and e-en- with a multimeter there is no connection but the plasma fires when started. Maybe somebody has an idea ? Do I need an additional relays ? With the hypertherm powermax 1000 G3 on the same mesa-card this didn't happen ?

Maybe sb has an Idea?

I would first verify that you have the correct firmware installed
in the 7I92T (7i92t_7i76x2d).

Next, I would verify the hardware by setting the 7I76 index input
mode to TTL mode (W4 left), unplug the encoder, then watch
the signal hm2_7i92.0.encoder.00.input-index in halshow
and check that it changes state when you ground the IDX+ input
(TB3 pin 13)

Hi,
I am new here, just rebuild an existing router and swapped the old electronics out for a mesa 7i96s with linuxcnc 2.9.4. everything went smooth except one thing.
The old router configuration had a single home sensor for a dual gantry. So i added an extra sensor for the second driver to be able to home. The problem is that it is impossible for me to add the sensor at the exact right spot (both are on opposite side of the machine.) So i managed to get the sensors roughly at the same spot (within 1mm) Now i would want to put the gantry manually square before powering up (it is a very sturdy bridge which automatically squares out if run by hand) and then i would like to power up the machine and jog incrementally up to the sensors. with the hal watch i could see which one lights up first and i could also check how much extra jogs i have to make to light up the other sensor. this would give me the offset i need to square out the bridge with homing. Now i am forcing the gantry out of square with the error in my sensor placement
SO here my queation: Is there a way to force linuxcnc to jog both joints simultaneous before homing

I'm beginning to wonder if it's getting speed signal or worse this encoder can't read in reverse. If it doesn't get speed feed back, it has nothing to hold it to so it just defaults to a set number. At least this is my reasoning thus far.

Hi Guys,Recently I had my 5i25 card fail and bought a 7i92TM to replace it, updating my HAL file to suit the new ethernet board.
LinuxCNC runs as normal, however, the spindle index pulse is no longer being recognised. The index pulse was detected and rigid tapping worked when I had th 5i25 card installed. I have not made any changes to the HAL file in regards to the encoder.

I confirmed with an oscilloscope the encoder is outputting the index pulse.
Halscope will detect the A and B pulses but not the index pulse.
The index pulse is not detected in either LinuxCNC 2.8.0 nor 2.9.4 (rigid tapping just hangs)

Is there something I am missing in my HAL file or is there a timming error in which the encoder pulse is not being read by the 7i92tm? 

Thanks a ton to all of you: jjdege, Billiiles !!!

Off course your hints were pointing into the right direction. I just had to go further than I thought I had to.
In short:
- added debounce
- increased debounce.0.delay

That was it!

I have manual tool change, prompting me for the tool required, hit OK, after change and then it measures tool length automatically. So happy!

Along the welcomed outcome, I stumbled across several problems. I'll wrap up my notes and write a comprehensive howto in due time.

Greez
a so happy chris