2

I am working on a CNC application where one controller board will be sending signals to another board. The environment may have electromagnetic interference (EMI) from nearby motor activity, so these relatively low voltage (3-10V) signals may experience corruption if not using shielded wire.

When transmitting signals from one PCB to another using a shielded wire, is it intended that the uninsulated wire (shown as #5 in the figure) be used as ground reference, OR, should a separate insulated wire be used?

My assumption is that the foil is a mylar film, with a conductive layer inside, and that the uninsulated stranded wire is there to ensure that this mylar foil is a continuous conductor along the length of the wire, forming a faraday cage around my conductive signal wires.

My understanding of the principle of a faraday cage is that when electromagnetic waves pass through it, current is produced and that this current both constructively interferes with itself and gradually drains out to ground.

SO MY QUESTION: is there any reason why I should not use the faraday cage as a conductor for the ground reference (every voltage or signal must have some reference potential)? OR is this something we avoid so that the faraday cage can do its job?

user391339
  • 273
  • 1
  • 7
  • I would assume the foil from the cable and 5 are shorted out, did you try testing it with a meter? I could be totally wrong, but I think the reason why 5 exists is because otherwise it would be awkward to connect the foil itself to ground. – Evan Carroll Feb 18 '22 at 06:27
  • Yes they are, and it seems that the foil is a conductor inside but not outside i.e. it's metalized on the internal wire wrapping side – user391339 Feb 18 '22 at 06:28
  • So the answer then is obvious (if I understand this right) if the other side of 5 is grounded, you can treat it as a reference to ground. I mean the ground wire in your house is one gigantic long wire subject to all that complex math stuff that I don't understand. But in a lump-circuit model, it's all just ground you don't worry about that for such short runs. The foil is shorted to 5, 5 is grounded (not an antenna). 5 can be reference ground in your circuit. – Evan Carroll Feb 18 '22 at 06:32
  • It could be the case, but I am trying to prevent interference from corrupting signal travelling on a wire from one system to another. Intuitively I want to keep any interference or induced currents/potentials away from my signal using this shielding, directing it to ground. This could be an argument for having a separate insulated conductor as my signal ground; in theory it would never see the EMI. – user391339 Feb 18 '22 at 06:38
  • This is way out of my paygrade, but some devices are ground-isolating for just this reason. You may want to ask when that's necessary. It's never been a problem for me. ;) – Evan Carroll Feb 18 '22 at 06:43
  • Short answer: no. The uninsulated wire is for connecting the circuit to the shield. How shield connects to your circuit's 0V line varies. If your module has a distinct "FE" connection like some PLC stuff, it would go there. Chassis lug too. Deciding whether, where (eg could be at other end of cable), and how (eg DC or AC via a small cap) to connect shields for EMC is an involved topic... see the EE stack – Pete W Feb 18 '22 at 13:58

0 Answers0