[etherlab-users] etherlab-users Digest, Vol 13, Issue 8

[Florian Pose]

On Tue, Jun 24, 2008 at 10:10:47PM +0900, Hobin Yoon wrote:
> Although you said the messages are good, it doesn't work as I
> expected. Inside the timer function, the out PDO doesn't make the
> slave work (blink LEDs) and in PDO values doesn't change over time.
> The correct behavior is verified by TwinCAT. That makes me think there
> might be some configuration that is automatically done in TwinCAT, but
> has to be done explicitly in EtherCAT. Doesn't that?

The configuration that has to be applied depends on the slave. If you
are not sure, which configuration is needed, please have a look at the
configuration applied by TwinCAT and use it in the same way. If that
doesn't help, please post the configuration (i. e. startup commands and
Sdos) from TwinCAT and your initialisation code for the EtherCAT master.

Without knowing the slave, it is very hard to guess the reasons for the
process data exchange failing.

> EL9820 uses Beckhoff FB1111-0142 and our in-house board uses Beckhoff
> FB1111-0140. According to the hardware data sheet
> (http://www.beckhoff.de/download/Document/EtherCAT/Development_products/FB1111-014X_V13.pdf),
> the former uses "32 bit In/Out digital interface" and the latter uses
> "16/8 bit asynchronous Microcontroller Interface". Are there any
> changes to be made to EtherLab because of the difference?

As long as the slave conforms to the EtherCAT standard, there should be
no changes necessary.

I noticed you are still using the EtherCAT master version 1.3.2. Please
try the master version 1.4.0-pre:

http://lists.etherlab.org/pipermail/etherlab-announce/2008-July/000007.html

-- 
Best regards,
Florian Pose

http://etherlab.org