Advanced Search

Upload the current files.
Just to be sure, that is a 4 axis/ 5 joints machine where Y axis is tandem?

To the left of the ribbon cable, are those optocouplers?
If yes, see what wires go to those optos inputs, those should be step, dir, and maybe enable, one is ground, and for testing you only need 1, the step pin, and if used the enable, but if the drives lock when powered on, then skip enable as they are enabled, so use a signal generator or parallel port, or a small crappy transformer that outputs 3 to 6V AC, should be enough to check if they work, and if they do, leave them on.
Be aware that some older drives use two inputs as step and have no dir at all, so steps on one pin turn the motor one direction, while steps on other pin move motor the other direction.
There are also "quadrature inputs" using steps on both pins, but 90 degrees between them, also no dir.
And as the last type, the 4 input ones, so when testing if the motor does not turn but just vibrates in place, check other pins, there should be 4 that cause the motor to vibrate.
All of those can be controlled by LinuxCNC, even through a parallel port, hence my hesitation to change the drives.
But if you think the time is more valuable, do change them.
And if the motors turn out to be 5 wire ones, they can be changed to bipolar, just require a bit of time and a soldering iron.
BTW, never ever remove the rotor from the stator on stepper motors, that will render them useless as they will de-magnetize.

Prototype boards arrived.

Wortwhile checking, you almost said it yourself.

You can also check "sudo dmesg" for suspicuous messages from the ethercat server. 
Provided your version logs diagnostics that way. The ethercat server installed with "sudo apt install linuxcnc-ethercat" does it anyway.

Ok, this must be the end of the road for this topic.  I'll try a new topic as I don't even see the A axis on the screen.
Anyhow, I want to thank everyone again for your help as it was a good learning experience for me.

Regards,

Routerworks

Yes , in 2.9.0-pre0 i use ec-debianize.
Now stay test debian 12 with ethercat and have execute step-by step procedure, writed by CORBETT.
Thanks for this.
I have see on debian 12 , but stay again study this problem , the value the all pin the Ethercat drives , they are written late in the associated hal pins, compared to the version I currently use with debian 10 and I can't understand why.
I tried using ethercat version 1.6 and 1.5, but the problem persists.
basically, once linuxcnc is started, the position values ​​of the axes are written in hal after a few seconds and starting from the sequence order of connection of the drives, that is, first the position value of x appears, after 1 second y, after z and so on, after having finished writing all the drives the positions are updated quickly.
Could it be that in the xml file, I configured drives that in reality are not on the machine?
I do this so as to still have all the associated pins in hal and then depending on the machine configuration they work or not equally.

The asm1083 is a PCIe to PCI bridge. The other chip would be the chip with the parallel etc functions.
Photos are ok, but the real bit of important info is the PCI ID number.

If you look at the text file I attached a couple of posts ago as well as having info regarding the parallel port cards it also has a couple of commands as where to get the info from the file system. Like what mode the card supports and stuff like that.

Any update?

Congratulations on the development!!!!!

I was able to test using the files, in version 2.10..

I have a machine here with ethercat and I can get 30mxmin, I would like to put it on it to test...

I tried to copy only the .so file in the modules folder, but I'm sure that this is not the correct way...

How do I make it work in version 2.94 normally???

 

Basic setup means C6930 from beckhoff master with native ethercat drivers which control beckhoff el6730 profibus master in the equidistat mode which control Metronix ars servopacks single phase power which control different type of servo motors like custom feedback SZGH 2kw + resolver etc ... For configuration I typed a software which will generate .xml file automaticly which I need to publish.. 

But now I am thinking about multi ethercat master something like instances to communicate over more ports ... Because it has 2ms scan time and other stuff I am cycling in the 250us

For this purpose the servo drives ARS need external counted spline acceleration curve stuff.. and velocity command too..

Which I can not code myself because of lack of experiences, knowledge and time too...

Thanks for link, I will check them all

Thank for your job.. it is brilliant !

I don't think they are, but I'll check, I think the ones from each driver go to the ``estop'' system of the machine or to the end switches, something like that... the signals are in this flat ribbon cable, but who knows what they are, maybe some older control system, some ttl something, I wouldn't know how to check that without schematic , and I couldn't find a schematic for this driver, the controller is from the 70s.

OK, thank you, i am a bit fuzzy today on everything due to lack of sleep and some other "issues".

Hi Endian,

link

what and which should be declared for basic workbench setup
After install, you can run it. See link how to setup and run.

 

The 3 bottom wires on the drive boards, what do they have written ?
Do you have an old PC with parallel port?
If those wires are step/dir and GND, fire up the PC, wire some wires from parallel port to those 3 wires, move the machine!

tommylight wrote:

Do they show driver and module in use?
I would guess at yes... as the flags are only "medium devsel"

No, neither of them. The last picture in each post is everything coming out of sudo lspci -v

It seems it has something to do with Sunix chipsets.