Ethercat HAL driver

Hello @bkt

Seems I'm a very lucky man. ....... Thanks for the advice and I hope in your or Chimeno help tomorrow.

Good Night

He does not answer you because the user @db1981 knows a lot more than I do, if someone can help you it is him.

Greeting
Chimeno

I see I explain me bad.

I started working with linuxcnc-ethercat installation in May .... in June I started my motors and let them go on the test bench, without gearboxes or mechanical loads, at the end of August. So hundreds of hours of system testing. No anomaly detected. Now that I have the motors installed on the machine I see the same thing so it goes without saying that everything works as you say and I made sure my cpu and my ethernet port held everything up.

Normal usage: These is all ok.

What happens: every now and then, when I turn on the machine or when I re-turn on linuxcnc after closing linuxcnc but not the machine, only on these occasions and never during work .... the drives do not go to OP. Yes, sure for initial communication anomalies ... in fact sometimes drive 0 does not go to OP, sometimes 1, sometimes none. The elxxx modules always work correctly (apart from the episode with el5151 which, however, we said to be my cause, I took it used and did not know it was better to reset). What should I do in these cases:

I turn off the power of the drives and turn them off (I remove the connectors from the drives).

This way when I power up the drives they go into INIT. When I start linuxcnc the drives go to SAFEOP then PREOP and when I press the enable button they finally go to OP so everything works fine.

Now I think this depends on communication problems (grounding ?, spike? ... the pc is powered in 24Vdc with filtered power supply .... I don't think so .... bad filters in the drives? Maybe .... rather it will be due to excessive delay of the ethercat line at power up due to a non-optimal hardware ....).

in any case I was wondering if it was possible to avoid disconnecting the power from the drives and performing the same procedure via software, then putting the drives first in INIT (so they are forced to start a new connection and reset the buffers) and then re-initialize the ignition of linuxcnc ..... if there is a way to always put the drives in INIT before starting linuxcnc it is fine. Can I use a script or a C ++ program to do this I ask.

Very thanks for all.
bkt

chimeno wrote: Hello @bkt

He does not answer you because the user @db1981 knows a lot more than I do, if someone can help you it is him.

Greeting
Chimeno

Very thanks for reply ... I write your name only because db1981 write " I reply you tomorrow xxxxx" and wrote your name in the post.
I didn't want to pull anyone by the jacket. It was a form of respect.

I may not have made it very clear that I greatly appreciated the help received from db1981.
I'm just sorry to have bothered for such a stupid thing.

Very thanks
bkt

Hello bkt,

for clearify : with OP we mean ethercat state OP. (not the drives power enable)

It should been like this:
after you start linuxcnc, all ethercat slaves must be in OP State (our system is running, communcation is established) , with closing linuxcnc the slaves falls back to preop state. Then you can start linuxcnc again and all slaves has to go to op state again.

Linuxcnc Estop or Enable button has no influence on the ethercat state machine!

Pleasy verify this again...
What servodrives are this exactly and is there a manual available?

You are in right ...... I'm sorry I didn't notice this ..... so the problem is that once Linuxcnc is turned off the drives go into PREOP. This is OK. But sometime when I start Linuxcnc they don't return to OP.

The problem is this.

I confirm that estop and other signals work as you said.

The drives are Maxine Ep3.

After this problem has appeared I can reset everything just by putting the drives in init or by turning off the power supply of the drives and after a few seconds turning on again.

bkt

best way would be to find out why they don't go to op again.
Next time this is, check the dmesg log. There must be ethercat information. unable to assign pdo, unable to set dc conf etc....