Ethercat AX5206 stuck in SAFEOP

I am having trouble getting a Beckhoff AX5206 to transition into OP mode.

It is a dual channel servo drive and has two servo motors attached to it. In twincat it configures properly into the OP state. I am sending the identical SDO values to the drive under linuxcnc ethercat driver which works except it stays in SAFEOP instead of transitioning to OP.

The ethercat lcec driver starts properly and all of the PDO mappings are working appropriately. No errors are stored in the drive. I query all of the "error" related IDN's and they remain empty.

'S-0-0021 IDN-list of invalid operation data for CP2 (PreOp)' - empty
'S-0-0022 IDN-list of invalid operation data for CP3 (SafeOp)' - empty

On the same machine I also have 2 AX5125 drives. They are configured properly and go direclty into OP mode correctly.

Has anyone run into this issue? I saw somewhere in etherlab-users mailing list that someone had run into the drive not going into OP when they would configured both drives but it would go into OP mode when they configured only one of the drives (it didn't matter which) but there was no resolution posited for the issue.

in syslog - Ethercat is giving me an error on slave 8. timeout while setting state SAFEOP, but the drive is in SAFEOP state.

Any recomendations or ideas?

Hi cman1024, I use the BECKHOFF drivers of the AX5000 series and the truth is that they are a bit difficult to configure and usually fail also in the loss of synchronism, it seems to me that the problem you have is with one of the PDO? It could be that some configuration is wrong, the first thing you would do would be to configure only the driver first. I attached it to you
Regards
Chimeno

I think that the drive is taking too long to transition from PREOP to SAFEOP.

In syslog I found this:
EtherCAT ERROR 0-8: Timeout while setting state SAFEOP.

but the drive does ultimately end up in SAFEOP and has no errors.

It does take quite a few seconds to write all of the PDO's to that drive. Either I need to speed that process up, or increase the timeout of the ethercat component while it is waiting for the drive to change to SAFEOP.

Is there some setting that I could change? I haven't been able to find it.

Hi cman1024,
As far as I know, it is only necessary more time in the AX5000x200 series, currently linuxcnc-ethercat expects each device to be in OP mode before moving on to the next one, which works in TWINCAT3 does not necessarily have to work in linuxcnc-ethercat, I am sorry not to be muya help, try to check the configuration that is all correct.

Regards
Chimeno

This is the change of must change in the igh ethertlab master
+#define EC_AL_STATE_CHANGE_TIMEOUT 15

Hi cman1024,
I would like to know what configuration igh, debian, linuxcnc-ethercat, linuxcnc, applying the PLL patches of these using for the AX5000,

Ok so I have solved this problem. The dual channel AX5000 does take a very long time to transition from PREOP->SAFEOP->OP.

I changed this in the ethercat master to 10 seconds.
+#define EC_AL_STATE_CHANGE_TIMEOUT 10

After that things have worked reliably.


I started out with the basic config that you gave me and that was able to get my drive to transition to OP. As I added more and more of the SDO config back in, it started to get back into the timeout error. Once I changed the ethercat driver to have a 10 second timeout, then things worked.

Thanks so much for your help!