Advanced Search

Is this possibly Plasma/EMI related?

That is, do you still get communication loss when the Plasma source is off?

andypugh wrote:

I can't help with a fix, but it might help to work out where the fix belongs.
Try plotting the position-cmd in Halscope to see if the problem is in the commands from the WHB or is inside HAL itself.

I'm attaching a short video and my configuration files. I've been trying to use Halscope but I still don't quite understand it. Tomorrow with more time I'll give it another try.

youtube.com/shorts/L8vz-2Y_pZM?feature=share

  I ran pnconf wizard and made a configuration for my CNC plasma table. I am able to home and jog my machine for about a minute until I get this error. 
It seems like the ethernet connection is disconnecting on its own somehow. Any ideas on how I could get the ethernet to stay connected so I can use my machine.

I have a probe that bounces during the G0 move away after probing.

I replaced the optocoupler, and the output is working again.I attempted to set up the tool changer. Unfortunately, I keep receiving the following error messages when starting LinuxCNC. Attached is my .ini file.

I don't think that a better debounce is the solution. We really ought to be able to solve this in the base code. Unless you have a probe that ounces after the machine has stopped moving.

You may have set the deceleration and acceleration ramps too steeply 

The controller doesn't care about the size of the motors, and there are many working EtherCAT systems out there running LinuxCNC.

You probably want to be running the master / dev version of LinuxCNC at the moment as there are various new features aimed at EtherCAT in there (for example the ability to run a initialisation step in the drivers as soon as realtime starts).

First see if you can work out or find documented the Denavit-Hartenberg parameters for the robot.
Also look at the sample robot configs which include a visualisation of the robot, For example:
linuxcnc/configs/sim/axis/vismach/puma/ in the config picker.

Have you tried creating a config with either the built-in PnCConf wizard, or the MesaCT wizard?

linuxcnc.org/docs/stable/html/config/pncconf.html

github.com/jethornton/mesact

I can't help with a fix, but it might help to work out where the fix belongs.
Try plotting the position-cmd in Halscope to see if the problem is in the commands from the WHB or is inside HAL itself.

If you can measure the pitch then you might be able to find it here:
connectorbook.com/identification.html?N=..._full_size_rect_conn

But it looks like standard 0.1" pitch so many things might fit. I would try the "dupont" style.
connectorbook.com/identification.html?N=&n=shroudless_plugs

On the top Right menu of the boot screen I have options for: 
Default Xsession
Gnome classic on Xorg
Gnome Classic
Gnome on wayland
Xfce Session
Only the Default Xsession and the Xfce session works. The others give me a white screen that says "Oh no. Something has gone wrong and the system cannot recover. Please log out and try again". There is also a sweet little icon of a computer screen and a sad face. Much of the googling I have done suggests that onboard only works with a few window managers.
Thanks again

Got the inputs figured out.

1.  Program drive so all inputs are 'no function assigned'

2. Change xml input address HAL name:
3. Add bitslice to HAL, using the bit chart from the servo manual:
Now I can press buttons connected to the drive and all five inputs react in halshow.  Rather exciting...

Next step is to see if I can fumble through changing the PDOs to 'fixed' and somehow map the outputs.

Anyone have a clue what 'sub-index 1' means?