[etherlab-users] Beckhoff AX200/Danaher S600

Florian Pose fp at igh-essen.com
Thu Oct 30 10:07:38 CET 2008


Hi,

On Tue, Oct 28, 2008 at 05:43:15PM +0100, Jens Kotlarski wrote:
> EtherCAT: Scanning bus.
> EtherCAT ERROR: Sdo upload 0x1702:00 aborted on slave 0.
> EtherCAT ERROR: Sdo abort message 0x06020000: "This object does not exist in
> the object directory".
> EtherCAT ERROR: Failed to read number of mapped Pdo entries.

this is an incomplete implementation on the slave's side. 0x1700 for
example is one of the indices of the assigned Pdos. If a Pdo with this
index is existing, the corresponding Sdo with the same index must exist
in order to let the master read out the Pdo's entries and their bit
sizes.

But this is no problem for you, because both the master and the slave
'know' the correct mapping for that Pdo. The slave knows it anyway,
because it is fix and the master knows it because the application (your
EtherLab model) told it of the Pdo entries, so it can calculate the
process data image correctly. The only point is, that the master is not
able to check, if the mapping is correctly configured. But we all know,
that it is like the application expects it, so you can ignore that for
now. But it is not very nice. You should tell Danaher, that this is a
non-conforming behaviour: Even if the Pdo mapping is not configurable,
the appropriate Sdos must be existing.

> We tried to scan the sent packages using wireshark. Therefore, the device
> ec0 has to be set to "up"... But, it seems that the device ec0 is not
> created (it cannot be found in /dev). Just EtherCAT0 is created when
> ethercat is started (which cannot be set to "up"). Is it missing or is it
> just renamed to EtherCAT0? 

There is no network interface 'ec0'. This is only the name of the device
for logging purposes. You can indeed configure the master with
--enable-debug-if to get 'virtual' network interfaces to connect
wireshark to, but this will not work from RTAI context. You'll have to
connect a switch between master and slaves to 'listen' to the traffic.

BTW, if fixing the control byte won't work, please send me the logs of
the configuration with 'ethercat debug 1' enabled.

-- 
Best regards,
Florian

http://etherlab.org



More information about the Etherlab-users mailing list