Advanced Search

I have a solid state relay OMRON G3PE 525B, for what I know it will not need flywheels

I forgot to mention this happen only if the first slave is connected (EK1100 --> First slave --> second slave --> etc... )
if EK1100 --> disconnect --> second slave --> etc... it doesn't blackout.

so my guess is that it's not an electrical or EMI issue, but something software side

Does that relay have a diode in parallel with the coil?
Mind the polarity.
Relays are inductors and as such create a lot of interference when switched on and off.

I don't know exactly what happen inside linuxcnc but I'm having this issue :  Ethercat issue video
at LCNC startup I power on the servos via a relay and the the whole Beckhoff modules blackout in LCNC but stays on physically, the lights don't even flash. 
This causes the first and second servos (sometimes the third) to not be set as operative.
If you want to try out yours I had to setup on HAL this to turn on halshow at startup: loadusr halshow
Any guess?

I don't get it. I meant they removed the software from the product page, now it's elsewhere..
If you mean they are crap quality, wich ethercat servos you suggest at this price point?

replace x with your servo/spindle index

net s-vel-cmd <= spindle.0.speed-out
net s-vel-cmd => cia402.x.velocity-cmd

You will need csp mode 0 :

setp cia402.5.csp-mode 0
setp cia402.5.pos-scale 1

Not a lot to go on. Where are you assigning a value to what parameter, where are you trying to access it and what happens when you do?
Usually best to give a short example of what you are trying to do and describe what is not working as expected.

I have tryed both, unfortunately #!/usr/bin/env bash makes no difference. Also halcmd show signal s-index-zero-for-ATC stays false. I also tryed to call M100 from a remapped M19 but it's totally ignored, meanwhile the same sintax for similar commands like M10 or M21 works fine.

UPDATE: somehow in toolchange.ngc M1xx doesn't work, just hang the M6Tx called BUT I noticed it works inside the subroutines called by toolchange.ngc, like M10 (remapped) so I temporarly solved.  What it could be the cause?

Hi NTULINUX,

I finally had the chance to try out your Gentoo-based LinuxCNC distro on my N100 system — thanks for putting it together!

At first, I was a bit surprised (and honestly a little disappointed) because the initial latency results were significantly worse compared to my Debian 13 / KDE setup. I even posted my findings yesterday to share what I saw. However, after digging a bit deeper today, I discovered that when I installed your Gentoo distro onto a new NVMe drive, the BIOS settings had somehow been reset to defaults. That definitely explained the poor initial performance!
Once I re-applied the correct BIOS settings, the latency results improved considerably. While it's still not quite as good as what I get with my Debian 13 setup, it's much closer now. This seems to reinforce just how well LinuxCNC performs on Debian 13 — especially considering that my Debian system has full driver support, including for sound, whereas your Gentoo distro doesn’t include those drivers by default. When I disable the sound drivers on Debian, I get even lower latency, though honestly the out-of-the-box performance is already excellent, so no need to disable them.

I also noticed one strange behavior on Gentoo: although the Latency Histogram window is otherwise responsive, the graph and Elapsed Time sometimes appear to freeze and stop updating. Pressing the Reset button restarts the process, which confirms that the application itself isn't hanging — just the display updates seem to stop intermittently.

Below are the latency graphs for both systems after running for 30 minutes under the same conditions:

• 5x glxgears
• Heavy internet browsing with YouTube videos playing
• System stress via:
  

  $ stress-ng --cpu 4 --iomix 2 --vm 1 --vm-bytes 1G --timeout 2000s --metrics-brief

  
  
  
  
  
  
  
  
  Thanks again for your work on the Gentoo distro — curious to hear your thoughts!

<duplicate message; deleted>

Best bet is to google for the various internet resources for setting up vnc.
You may have some luck searching the forums, but it wasn't a package I thought needed, still don't, for running a CNC machine. I'm not a fan of running a CNC setup remotely.
If you just need to run an editor or even Linuxcnc remotely it can be via SSH.

When I think of it, it may not work very well with FF1 in this situation.
I guess orient just gives a position, without any ramping up and down.
Not like a motion profile with acceleration up and down.
My new bet is to set all FFs to zero and slowly increase P.

I´m using a Mesa 7i33 for the encoders and pwms. The encoders are Fanuc, so there is very little info to get on the specs of the encoders. All I know is they are TTL and 2000i/rev
All I have additionally is the schematics of the control where the encoder connection can be seen:
 

As it shows inverted channels I think they are differential types
And yes, the 7i33 is jumperd to TTL (Jumpers DOWN)

Hello, I am writing G-code files with parameters O<measure> call [#1] [#2] ... and I cannot access certain variables in LinuxCNC 2.9.4 (e.g., <#_tool_diameter>). I need to read this data in my tool table in order to develop my program. Can anyone help me?Thank you.

I am sometimes facing the same error just when the probe is tripped during the slow probe move. But I don't fully understand how this is related to the probe home?

Check my comment on page 9 for a csv config.