[etherlab-users] EL6751: invalid input configuration, wrong physical address of sync manager

[Stefan Laufmann]

I had the same feeling. But I do not understand why TwinCAT chooses to use
other addresses for the PDO sync managers. Nevermind I tried to change the
addresses in the TwinCAT configuration to the ones that Etherlab uses and it
worked just fine.
Therefore I think that the address and length settings of the sync managers
are not the problem but the point that Etherlab wants to set them as
disabled. I have attached a dump of dmesg from the run.

I assume that Etherlab disables the sync managers because I don’t configure
any PDOs which is because I don’t know the Interface the EL6751 will offer
after the configuration.

Greetings
Stefan Laufmann
-----------------------------------------------------

Stefan Laufmann

ITPower Solutions GmbH
Kolonnenstraße 26
10829 Berlin

Tel.: +49 (0)30 6098501-20
Fax: +49 (0)30 46799891
E-Mail: Stefan.Laufmann at itpower.de
Web: www.itpower.de

> -----Ursprüngliche Nachricht-----
> Von: Gavin Lambert [mailto:gavinl at compacsort.com]
> Gesendet: Mittwoch, 27. Mai 2015 01:42
> An: Stefan Laufmann; etherlab-users at etherlab.org
> Betreff: RE: EL6751: invalid input configuration, wrong physical
> address of sync manager
> 
> On 27 May 2015 04:01, quoth Stefan Laufmann:
> > I am still working on a running configuration for the EL6751 clamp.
> > After analyzing and comparing the network traffic of both TwinCAT and
> > EtherLab I finally understood what is causing the problem.
> > At least I think I do.
> > After sending the startup SDOs and getting the EL6751 slave into
> PREOP
> > mode the sync managers for buffered access (the PDOs) are configured.
> > But TwinCAT sets them to the start addresses 0x1800
> > (Write) and 0x2400 (Read) whereas EtherLab sets 0x1000 (Write) and
> > 0x1400 (Read).
> >
> > Is there a way to influence the addresses for the sync managers out
> of
> > a kernel module that uses the functions provided by "ecrt.h"?
> 
> As far as I am aware, Etherlab just uses the configuration that the
> slave defines itself.  (From the SII for the mailboxes and from the
> slave's own internal initialisation for the PDOs.)
> 

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: log_2015-05-21_rev0x13.txt
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20150528/5241d42f/attachment-0005.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3056 bytes
Desc: not available
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20150528/5241d42f/attachment-0005.bin>