NJ EtherCAT shielding.

[Andy_P]

With regard to an NJ EtherCAT network, all comms cables are shielded, but where should the shield connect to ground? I don't measure any connection between PE or FG on the NJ's power supply to the outer casing of the EtherCAT port on the NJ. (No connection to the Ethernet/IP port either.) I am also using a GX-JC06 junction slave, and none of the outer casings on this unit connect to ground either. On Omron G5 servo controllers, the EtherCAT link in and out port casings do not connect to ground either, but they ARE connected together, which means that the shielding path is maintained through. Is the grounding somehow made through active components, instead of a direct passive connection? Edited March 25, 2015 by Andy_P

[ParaffinPower]

Ethernet outlets (Phoenix sell them) do this job. Depending if the EtherCAT nodes are local or remote determines if the shield is grounded or not (there's a plug you can remove to insulate from the DIN rail). I had long cable ecat runs to remote I/O recently, and these things were a silver bullet. Pp

[Andy_P]

Interesting, So in your case on the long cable run, did you place one of these outlets at the end and ground it? How would you typically ground the shielding in an Omron NJ EtherCAT system? Also, have you (or anyone else) ever used one of Omron's GX-JC03/6 EtherCAT junction slaves? On further investigation, the outer casing of the Ethernet ports on these units (and hence the connection to shield) is earthed via a 1 megohm resistor. This seems strange to me. Why would they not be earthed directly? Edited March 25, 2015 by Andy_P

[ParaffinPower]

Hi Andy I grounded at one end only. The remote panels were a good distance away, so ground may well have been at a different potential. These outlets are more suited to terminating installation cable (solid cores). They may have something suitable for in-panel use. Just be sure to get a 360 degree wrap on the shield, and no shield pigtails! Pp