Advanced Search

That's what makes it so challenging with ethercat. Settings in several places and cross-dependencies.
Anyways, the second line in ethercat-conf.xml has the appTimePeriod setting.
There also


There are two more settings you can do now when things start to work.
Add this line right under the "<slave" line

and go back up to the second line in the file and add a minus-sign to the refClockSyncCycles value

You are right.

Switched the 7I85 channel to TTL mode, disconnected A−/B−/Z−, and kept +5, A+, B+, Z+, GND only.
The encoders now start in IDLE every time and looks work.

These Chinese “RS-422” linear scales most likely don’t have a true RS-422 line driver and cannot drive the 120 Ω termination on the Mesa inputs.

As a side note — it seems these low-cost Chinese sensors are often mislabeled:
one of my units was even shipped with 1 µm resolution instead of 5 µm as specified.
So if anyone else is buying similar “RS-422” magnetic scales, be cautious — the specs are not always accurate.

This is a new version of the 'vismach' simulation tool based on VTK instead of OpenGL:
github.com/Sigma1912/vtk-vismach

Still work in progress.
For more information see the README and the comments in the files.
 

I had a pretty hard time finding a pc with a parallel port. I don’t use a mesa card

The new pc is a dell optiplex 980 That is a very compact unit. The old HDD doesn’t seem to fit in the housing.

I use pcloud for gcode synchronization and I just couldn’t open the site on 2.8. So I decided to install te latest version of LinuxCNC

Yes that is right.
Interesting that you mentioned stmbl, I remember they were integrated with some robot (video of manutec carrying a coin).
Do they send its position over smart serial ?
I suppose when position is requested drives can’t replay all at same time because of other priorities, but stll they have to provide position.
I have custom drives build and they are reading encoders and do control but I am looking next part integrating with kinematics module.

Hi Guys,

A plasma machine I built that runs the original PlasmaC is having a fatal error message that is stopping the machine from operating. 

   

The guys who run the machine tell me the machine ran fine last Friday and when they came to use it on Monday morning they got this error message not matter what gcode program they tried to run.

This same issue happened back in March, and the only way I could fix the issue was to reformat the machine to a default state. Since it is re-occuring, there must be something going on to cause the issue.

Ive attached the gcode file to this post. 

I finally got around to getting everything wired up enough to test.
At first, I couldn't get it to start using a relay and a power supply on a switch. I realized the manual says "or dry contact closer", so I wired the relay directly to start- pin 15 and start+ pin 34, which worked.

Now I can't get my arc ok signal. I followed the example wiring in the manual. I connected ark ok+ pin 23 to the relay. Then 24vdc+ pin 37 to the relay. Then ark ok- pin 4 to power ground pin 36.
Wired up like this, the arc will not start and the plasma cutter gives an error. If I disconnect the ground to ark ok- pin 4, the plasma will start without errors. I am getting 24 volts between power and ground, and the 2 ark ok pins will close the circuit when the arc starts.

I am attaching the examples from the manual plus diagram of how I wired the arc ok relay.
I may have wired it up like example 4(which it says don't use) a long time ago before I decided to remove the old interface connection on it, so maybe that messed it up.

Does my ark ok wiring seem right? Should I try wiring it differently or should I consider using a reed relay instead?
 

Poke. Figured I should stick this here.

Machine is back together with some new upgraded, ethercat is working for the xyz.

Now I'm on to making the spindle work with modbus.
Is there a simple step by step instruction to set this up? I got a usb adapter, and I'm running qtdragon. In addition to just tuning on and off, id like it to display the true spindle speed and the load.

Thanks

I thought so but obviously something is wrong. I did post my ini file a post or so ago. I was trying to change the direction the axis moves after the index is seen. Nothing I change seems to do anything

Do you ini file settings match the requirements for index only homing?

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

The STMBL drives can work with Yaskawa encoders.
(But supply is an issue)
forum.linuxcnc.org/27-driver-boards/5375...tmbl?start=30#311767
I see that you already know about their existence, but it might be worth considering using them.

Hi,
as I see from list there is no support for this type of encoder.
It's yaskawa absolute 12bit encoders. After power up some mS and it sends its absolute position over UART, after that if it moves it starts to send only quadrature pulses A/B like normal incremental encoder.

I would like to use the POSITION_FILE in combination with the index pulse
of the encoder to home my lathe. The idea is that you know approximately
where you are when starting up, and once you pass the first index pulse
you know the exact position. Has anyone done this... and if so can you
explain how?
 

Analogous to Cloudflare Pages automatic deployments from GitHub?

Thanks for your detailed answer!

If a PC running Linux breaks, try putting the old hard drive in the new computer.

It actually works perfectly 99% of the time!

Why did you install 2.9?

Which version of 2.9? 2.9.4?

How does your PC control the machine?