[etherlab-users] experience with CU1128 | EtherCAT junction

[Gavin Lambert]

I have one in a test system (though not using it in production) and it works
fine.  As Graeme said it basically just works without configuration, though
if you're specifying devices by network position then you need to be careful
about the ordering, since it will appear at three different places in the
network.

There's two aspects to Hot Connect.  One is the hardware auto-link and
network rescan that the slaves and the master do automatically.  The other
is configuring your application and network model accordingly.  If you want
to take advantage of it, you really should have at least the first slave in
each hot-connectable group have an alias address rather than a
position-based address and you should use separate domain groups for each
hot-connect group.  (Or alternatively don't rely on the domain WC state and
instead use some per-slave process data to determine if it's online or not.)
             
> -----Original Message-----
> From: Graeme Foot
> Sent: Wednesday, 19 April 2017 10:51
> To: Matthieu Bec <mbec at gmto.org>; etherlab-users at etherlab.org
> Subject: Re: [etherlab-users] experience with CU1128 | EtherCAT junction
> 
> Never used them, but just had a quick look at the doco (just cos I was
> curious).
> 
> - It mentions the ability to Hot Connect any of the ports, but there's no
> configuration behind this, just the usual hardware support.
> - It is detected as three ESC's (as that's what it has internally), each
one being
> identified by a different RevisionNumber (but same product code)
> 	- Primary ESC 1:	  ProductCode="#x04685432"
> RevisionNo="#x00010000"
> 	- Sub ESC 2:	  ProductCode="#x04685432"
> RevisionNo="#x00010001"
> 	- Sub ESC 3:	  ProductCode="#x04685432"
> RevisionNo="#x00010002"
> - It can be used as a distributed clock reference clock
> - There's no PDO or CoE data / configuration
> 
> So I would expect you just need to set up three slaves as above, then the
> master can report on the status of their ports (via reg information).
> 
> One thing to note is that TwinCat knows that the sub ESC's are sub devices
> but I don't think the EtherLab master supports this, so you may need to
place
> Sub ESC 2 and Sub ESC 3 further down your list of slaves.  (i.e. Sub ESC 2
after
> the devices connected to port 3, and Sub ESC 3 after the devices connected
> to port 5.)  See page 12 for a diagram on how the ports relates to the the
> three ESC's.
> 
> 
> Regards,
> Graeme.
> 
> 
> -----Original Message-----
> From: etherlab-users [mailto:etherlab-users-bounces at etherlab.org] On
> Behalf Of Matthieu Bec
> Sent: Wednesday, 19 April 2017 6:55 a.m.
> To: etherlab-users at etherlab.org
> Subject: [etherlab-users] experience with CU1128 | EtherCAT junction
> 
> Hi,
> 
> Has anyone had experience using a "CU1128 | EtherCAT junction" from
> Beckhoff (*) with the etherlab master?
> 
> It seemingly looks like a dumb device, but the documentation hints at
> particular steps to set it up with TwinCAT:
> 
> ```
> Topological configuration
> With the CU1128, special attention should be paid to the sequence of the
> EtherCAT slaves. Since the CU1128 has 7 junction ports, drop lines
connected
> to ports must and can be clearly identified in practice. If incorrect
information
> is provided in the configuration (TwinCAT System Manager file *.tsm), the
> system cannot start.
> ```
> 
> Regards,
> Matthieu
> 
> (*) https://download.beckhoff.com/download/document/io/infrastructure-
> components/cu1128en.pdf