Advanced Search

Ok, it seems r8168.aspm=0 or r8168.eee_enable=0 cannot be set with the grub customizer, but found another way:

sudo geany /etc/modprobe.d/r8168-dkms.conf

add the following:

# Disable Energy Efficient Ethernet
options r8168 eee_enabled=0

# Disable Active State Power Management (ASPM)
options r8168 aspm=0

Then save and close geany, then run the command:
sudo update initramfs -u

Then reboot
sudo reboot

After the reboot you can verify that the energy efficient ethernet is indeed disabled with
sudo ethtool --show-eee enp2s0
replace enp2s0 with the name of your Ethernet interface. You can get the name of your interface with the "ip a" command
If ethtool is not installed do:

sudo apt update
sudo apt install ethtool

I'm now running an endurance test with EEE disabled. Will report back here.

Yes. I have a real machine with QtDragon where this has been happening for some time now but I thought it was caused by something else until I noticed it happening in a simulation as well. I did some more experimenting and I discovered that if I have a G64 with a large P value, like 1 for example, the spindle doesn't restart. The smaller the P value is the further into a program it has to run before the spindle doesn't restart after an abort.With a G61 the spindle always restarts. I've only used 3D_Chips.ngc to test it.

I think this is related to what's happening:
github.com/LinuxCNC/linuxcnc/issues/579#issuecomment-640557582

This sure does sound related to #865 and #882, where readahead segments aren't dropped during task abort, and get flushed into the queue by commands following the abort: the ON_ABORT_COMMAND in this case, and the restored state tag in the case of #865.

However, in my quick test, I couldn't tickle the bug in the 2.8 branch by adding the ON_ABORT_COMMAND from @sheetcam's config. Anyway, if it is the same problem, hopefully #882 will fix it.

Hello,

i just had time to work on this:

i switched the old Thinclient with a fast new Ryzen 5600G PC and had worse problems than before even with gmoccapy.

I searched around the forum and found this:

Latency, error finishing read, and IRQ affinity - LinuxCNC

On the thinclient i had set up isolcpus already, but not irqbalance. Just don't know.
This + isolcpus=5,11 is now running on the new PC like in this link and the machine runs without errors on the first travels along the axes.

I havent't milled yet, but i guess it's working now.

Best regards,
minimal3d

 

first try
 

I do not see the issue with small programs. And TBH it sometimes does not happen unless I am a few moves into the program.. Once it does happen it will happen every time from any point - even as it makes its first move. Then if I change to a small program it seems to work fine. Change back to large program (been using the one posted here for testing) and the problem is back.

I will try to get the config zipped and posted here.

I will try and test 1 line at a time to see if I can narrow it down.

Ok, I got a setup running with 2 EtherCAT drives without problems. I've now added a third EtherCAT drive and the problem has returned, occurring about once every hour or so. I check the tmax write plus the tmax read for the lcec which combined to max 23.5% of my servo loop time.
I would really like to turn off the r8168.aspm=0 or r8168.eee_enable=0 functions to see if that helps as rodw suggested, but I do not know how. The grub customizer does not allow me to save those, throwing a "not found" error.

PCW wrote:

You might check the network latency with

sudo chrt 99 ping -i .001 -q [7i96s_ip_address]

where [7i96s_ip_address] is either 10.10.10.10 or 192.168.1.121

Let the command run for a couple of minutes and the hit control C
to stop the command and print the statistics.
 

Hi,

here is the result:

--- 10.10.10.10 ping statistics ---
216217 packets transmitted, 216217 received, 0% packet loss, time 216389ms
rtt min/avg/max/mdev = 0.044/0.061/9.253/0.062 ms

Seems there are no packet losses..

 

how many people want something like this?
i could build a small pyqt-gui that generates a small gcode preview using the 'rs274' tool.
but i already have a console tool that converts gcode to svg.

github.com/multigcs/riocore/blob/main/riocore/files/ngc2svg.py

This is the central system I configured when installing LinuxCNC on the image before. I'm not sure if it can be used

 HelloI haven't made any progress with EtherCat. The Beckhoff components are just not up and running yet. The LCNC GUI also doesn't seem to communicate with the underlying layer (CIA402). Why is that? a link helped me shed some light on it.
 it looks like everything is-fine.

 

but then something like this

 


Failed to register PDO entry


 

Installation of Linuxcnc Central Station
Install through the following steps:
git clone  gitlab.com/etherlab.org/ethercat.git
%Download EtherCAT Master source code
If git is not found, then:
This error indicates that it is not installed in your system. You can install it by following these steps: git
Installation: Git
Update package list: First, update the package list of your package manager:
sudo apt-get update
Install git: Use the following command to install: git
sudo apt-get install git
Verify installation: After installation is complete, verify if git is installed successfully:
git --version
If the installation is successful, you will see the version number
）
cd ethercat
%Enter the main directory of clone source code
./bootstrap 
%If the downloaded tar.bz package contains an m4 subdirectories, this step is not required; Git clone is necessary

In progress/ When bootstrap, the following error may be reported, as shown in Figure 1
  
At this point, you need to use the following command to install pkg config.
sudo apt-get install pkg-config
Continue with the installation.
./configure --prefix=/opt/etherlab --enable-8139too=no --enable-generic=yes
When executing again, if the result prompts an error, as shown in the figure below, you need to flip over to see where the 'no' appears, as shown in the figure below
    
 
To resolve the error, it is necessary to install gawk and make, using the following code.
sudo apt-get install gawk make
Continue installation.
make -j 4
%4 thread acceleration
sudo make modules
sudo make install
sudo make modules_install
sudo depmod
After installation, a configuration file is required. The method is as follows

cd /opt/etherlab
 
ip addr show
%Find the Link/Other option for the local network card and copy the MAC address of the network card
 
%Then edit the configuration file
sudo vim etc/sysconfig/ethercat
%Copy the MAC address to
MASTER0_DEVICE=“”
%Change VNet="" to
DEVICE_MODULES="generic"
 
sudo mkdir /etc/sysconfig
sudo cp etc/sysconfig/ethercat /etc/sysconfig/
sudo ln -s /opt/etherlab/etc/init.d/ethercat /etc/init.d/
sudo ln -s /opt/etherlab/include/*.h /usr/include/linuxcnc/
sudo ln -s /opt/etherlab/lib/lib* /usr/lib/

At this point, the configuration is basically complete. You can enter the following code to try starting the main station

sudo /etc/init.d/ethercat start
At this time, it will return

Starting EtherCAT master 1.6.0-rc1 done

Enter the following code to view the main station information

sudo ethercat master
If the prompt bash: ethercat: command not found, use the following method.
First, check if you are currently using bash or dash.

ls -l /bin/sh
If dash is displayed, use the following command to modify it to bash.

sudo dpkg-reconfigure dash

Then use the following code to modify the bashrc file.

sudo vim ~/.bashrc
Add the following code to it

alias sudo='sudo env PATH=$PATH'
PATH=$PATH:/opt/etherlab/bin
Execute source~/. bashrc
At this point, you can use the EtherCAT command-line tool.
  
Above, EtherCAT Master installation is complete (restart the computer for use).
 

.... 4 hour try to understand these .... my project linked on github ..... now It have these cmakelist (see attach) ... compile well ... and works .... tthe same cmakelist on original poject ... give me 147 of these error (like libtool was not linked)

:-1: error: /home/gc81/linuxcnc-source-dir/lib/librs274.so: undefined reference to `DISABLE_FEED_HOLD()'

some suggestion why one side it build  and works and other side not at all? ....

 

Hi Macwolf,
I appreciate your work and insights! I’m currently trying to connect EtherCAT with LinuxCNC using EtherCAT-master. I followed the steps outlined in the link below, but I’m running into an issue when attempting to load the lcec_conf file in LinuxCNC.

I receive the following error message:
./custom.hal:4 /home/cnc/linuxcnc-ethercat/src/lcec_conf exited without becoming ready

 As a beginner in troubleshooting these types of issues, I would greatly appreciate any guidance you can provide. If possible, could you also share the steps you followed?
Thank you!
forum.linuxcnc.org/ethercat/49771-etherc...ce-full-instructions     

I have not forgotten you, but have been too busy to test on my machine.

line 231 returns the tool to the start position, but with the new tool length.

52 #<local_start_z> = #<_z>
<snip>
probe
set tool table offset
return to Z location:
231 G0 Z[#<local_start_z>]

If the new tool is longer then the original, and the machine cannot go up anymore to get to that position, you would get that error.

you could add this just above line 231 so we can see where it is trying to go:
(DEBUG, moving to Z: %f#<local_start_z>)