Advanced Search

hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error queuing read! iter=22200)
hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error finishing write! iter=22200)
hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error queuing read! iter=22201)
hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error finishing write! iter=22201)
hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error queuing read! iter=22202)
hm2/hm2_7i96s.0: error queuing read! iter=22200)

hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error finishing write! iter=22202)
hm2_eth: ERROR: sending packet: Network is unreachable
hm2/hm2_7i96s.0: error queuing read! iter=22203)
rtapi_app: caught signal 11 - dumping core

Actually, I did have a problem with the closed loop ECT60 in the beginning.
First start it was ok, second start it gave me Following Error. 
So before linuxcnc starts the encoder position is reset to zero.
My start script
This may not be the right way to solve this, but it seems to work and I stopped researching the problem.
 

Thanks for Rod.

So those with only 4 cores should there be any issues ?

In case anyone is thinking the RPi may require at a bit of time cos of a library incompatibility that was seen earlier.

@Asymmetrisch
Ja, die liegen wohl am Anfang und in der Mitte, aber beim Darüberfahren kommt keine Rückmeldung.

@PCW
Yes, I know, that's how I had it originally, but there is no feedback from the index signal, so I commented it out again and am currently using the limit switch as a reference point.
It should actually be that when searching for the index point, the motor moves toward the limit switch, and if the index is not in the area traveled, it turns around at the limit switch and searches in the other direction.
The index should be at the beginning and in the middle.

Clear Linux is now EOL.

community.clearlinux.org/t/all-good-thin...clear-linux-os/10716

github.com/clearlinux/distribution

The 5.4.290 kernel debs are bugged, I was hitting that problem too with the screen going black. I need to downgrade the kernel and freeze the repo as there will be no more RTAI development. In the mean time, use the 5.4.279 kernel debs. 5.4.290 is broken. The latency test and everything else works, including the LinuxCNC sim, however when running an actual machine with 5.4.290, the system does indeed crash. I need to find the last working 5.4 kernel and make a few more changes to Kconfig, after that, RTAI will be considered frozen.

Debian Trixie is released Yeay!!

It contains Linuxcnc 2.9.4 Wow! 
To upgrade from Bookworm, very rough instructions

sudo apt update
edit /etc/apt/sources.list and replace any instance of bookworm with trixie
sudo apt update
sudo apt upgrade.
Check you are booting in to the new 6.12 Real time kernel
run sudo grub-customizer to set this.
Reboot into the 6.1.2 RT kernel

I removed my version of linuxcnc
sudo apt remove linuxcnc-uspace linuxcnc-uspace-dev

If it is a new install, download debian from Debian.org Trixie is their  current release and install with balena etcher

Check you are running Trixie: lsb_release -a
Check you are runing the prempt_rt kernel 6.12: uname -v (the RT kernel got installed for me)
Linuxcnc should install the RT kernel but I have not checked that.

Install linuxcnc. Seems one dependency did not play well

sudo apt install libqt5webengine5
sudo apt install linuxcnc-uspace linuxcnc-uspace-dev

enjoy Thats It. If you are running ethercat, I think the packages need to be removed and reinstalled as the config ssemed to be missing config files in 6.12

PS I am using the Cinamon Desktop, Latency seems good even without isolating any cores
Seems that I can run the RT kernel without this desktop hanging which is good! I'll see how it goes.
 
Have fun

Bumping this, I seem to be having the same problem.  
My tool change routine is here:
github.com/Funkenjaeger/fj-lcnc-cfg/blob...files/toolchange.ngc 
(and the rest of my config is in the same repo)

My tool changes sometimes work when I do them manually, but seem to halt when they're part of an actual program involving tool changes.  Specifically, when running a real program, the GUI indicates that it stopped at line 63, which is the last line before the first G1 command.  That checks out, because physically the machine moves to the right Z height for tool change (in accordance with line 63) but then never slots the tool into the pocket (as commanded in line 64)

Most of these little ones are linux. They are "ok", but most are pretty limited in features and performance.

I have a DDCSV3.1 and its cool to have an all in one single box, but it doesn't come close to linuxcnc. Contouring motion is very poor and its limited to 250khz pulse rates.

Thank you very much for your effort
I will try to test the revised version in my simulation.
I will report at the latest when I am back at the machine, or when I am back from vacation.
Thanks again for your work.
Greetings Michael

To use index, you would need to uncomment these and similar lines

#net x-index-enable joint.0.index-enable <=> hm2_7i92.0.encoder.00.index-enable

and change:

HOME_USE_INDEX NO
to
HOME_USE_INDEX YES

And know where the physical index mark is on your scale as Asymmetrisch mentioned.

Wenn die init-LED (CR6 bei einem 7I96S) aufleuchtet, bedeutet dies, dass ein Fehler aufgetreten ist.

Der wahrscheinlichste Fehler ist ein Watchdog-Biss aufgrund von Latenzproblemen.

Welche Meldungen werden gedruckt, wenn Sie LinuxCNC über die Befehlszeile starten?

The world is so beautiful because of the good people and those we can love.
Master Hakan, thank you so, so much!

Agree, most are running a proprietary embedded system, but this one is an ARM-based Linux system with 3d gpu. I wonder if it's a port of LinuxCNC and or we can mod it

 

5I25/6I25 config for 7I85 on P3 (ext DB25), 7I74 on P2 (internal header)

Only 3 sserial channels on the 7I85 (port 0) because of space considerations
All 8 sserial channels on port 1 (7I74)