0
\$\begingroup\$

I am working on a device that uses EtherCAT. On one side, I have the standard EtherCAT Connector - an RJ45. On the other side, I would like to experiment with different connectors. For starters, I will try a 6 pin JST connector.

I was wondering if there would be any concerns in making this adapter? Essentially putting an EtherCAT bus over a JST connector.

Are there any tests I can do (using an oscilliscope, or even an arduino) to make sure data is being transmitted effectively via this adapter?

Thanks for the help

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Ethercat is effectively 100baseT Ethernet. Why do you want to play with connectors? \$\endgroup\$
    – Kartman
    Apr 26 at 2:55
  • \$\begingroup\$ Since it is Ethernet, it can be used with industrial grade Ethernet connectors rated for Ethernet use. However, if you are going to experiment with custom connectors, please note that Ethernet also requires certain specification from the cabling/wiring, which usually means the industrial cables with industrial connectors are just industrial grade CAT6 or similar. \$\endgroup\$
    – Justme
    Apr 26 at 5:10
  • \$\begingroup\$ What's the part number for the connector? Presumably it has internal magnetics? \$\endgroup\$
    – raaymaan
    Apr 26 at 10:17

1 Answer 1

0
\$\begingroup\$

As noted by @Kartman, EtherCAT at the physical layer is effectively 100MBit ethernet. I have had extensive experience plumbing 100MBit ethernet through robotic systems that operate unsealed in some really nasty environments and it's pretty reliable:

-we would use JST PA, SH and GH connectors, crimped by hand

-we would use 24awg standed wire twisted with a hand drill

Certainly not ideal!

Our ethernet links would work well until:

-multiple shorts to chassis occurred (single point was kind of okay)

-corrosion caused the connection resistance to increase

with regards to resistance: My general guideline was if a fully connected link measured less than 2 ohms across a pair on a multi-meter (usually 4 to 4.5 ohms indicated with lead resistance factored in) then it was good to go. Above that I would service the wiring and connectors.

As for diagnostics, if you can connect a linux running device direcly to the link in question, netstat -i will report the number of dropped packets:

https://www.cyberciti.biz/faq/linux-show-dropped-packets-per-interface-command/

And with two linux appliances iperf can be used to benchmark the link:

https://www.dell.com/support/kbdoc/en-ca/000139427/how-to-test-available-network-bandwidth-using-iperf

\$\endgroup\$
1
  • \$\begingroup\$ Thanks, this is really helpful advice! To clarify, we are expecting some packets dropped when going from the RJ45 to the JST connector? What I have is an RJ45 on a PCB with traces going to the JST connector. Thus, the resistance should be minimal. Do you forsee any issues with this approach? As for the Linux testing, is it simply the case that you send out a packet from the Linux device, and then read it again out of the system? It's essentially a loop? \$\endgroup\$ Apr 28 at 1:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.