Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Ethercat / xPC target

17 views
Skip to first unread message

Stuart Fallon

unread,
Sep 26, 2012, 7:04:06 AM9/26/12
to
I'm trying to set up a Simulink model (in 2012b) for Ethercat communications. I've managed to establish communications with the Beckhoff slaves (the LEDS are indicating OP mode and wireshark is showing Ethercat packets on the bus). The slaves are an EP4174 DAC and an EP3174 ADC and Im trying to send a test signal from the master to the DAC and then back to the master via the ADC.The EthercatInit block returns [0,0,8,5] but it is not documented what this vector is. I'm guessing that the 8 is the state of the slaves (?) and I can get [0,0,8,0] and [0,0,8,4] by playing around with various distrubuted clock settings on the ET9000 side. However, I've still had no success in stimulating the outputs of the slaves yet.

Any suggestions would be welcome.

Gordon Weast

unread,
Sep 26, 2012, 1:11:36 PM9/26/12
to
Stuart,

For a problem like this, you should be working through MathWorks tech
support.

If you haven't installed the update for bugreport 834038,
http://www.mathworks.com/support/bugreports/search_results?search_executed=1&keyword=&release_filter=Exists+in&release=224&selected_products=2679&commit=Search

Please get the attachment appropriate for R2012b and install it. The
newest ET9000 produces a config file that doesn't work with the older
version of EtherCAT that ships with R2012b and earlier.

If you're still having trouble after installing the patch, please
contact MathWorks tech support about it.

Gordon Weast
xPC Target Development
MathWorks

Simon Mushi

unread,
Nov 6, 2012, 12:32:08 PM11/6/12
to
Hello Stewart,

Have you been able to resolve your problem? I would be glad to compare notes as we are having the exact same issue with xPC target as an EtherCAT master communicating with a Beckhoff EP4374 ADC/DAC node. When we start our model, the node goes into OP mode (based on the LEDs) but all the Rx variables returned are zero.

However, when we used TwinCAT as master instead of xPC, we could see the correct Rx objects returned corresponding to a sine wave input from our function generator.

We also use TwinCAT to generate the ENIs.

Regards,

Simon

"Stuart Fallon" <w82...@googlemail.com> wrote in message <k3unf6
$apd$1...@newscl01ah.mathworks.com>...
0 new messages