0
\$\begingroup\$

Everywhere we look at, the same recommendation is made: "never run coaxial cable parallel to electrical cables". Well, I get it, but.... I have no other option™. More than that, those cables have to run inside the very same conduit (about 1" thick) for less than 1m.

Well... the same guidance is given about ethernet cables. And I never had problem with CAT-5 cable alongside electrical power for about 1 meter length (maybe that ethernet section cannot reach 1Gbps speeds, but I don't need that either).

My question is about quantification of this "never run [...]". That is:

How much signal degradation (or any other harmful interference) should I expect in a coaxial cable 75ohm in parallel to 127V AC @ 60Hz for some given length?

An approximation is more than enough. How does it compare to other common sources of interference, say, CMB or airports nearby?

If current is relevant, consider that these electrical wires power a medium LCD TV set (like 120W in use, or 2W in standby).

If the nature of signal is relevant, the coaxial cable transmits 50mbps full duplex from the operator router ca. 100m away.

\$\endgroup\$
12
  • 1
    \$\begingroup\$ Not an answer, but you may benefit by adding ferrite sleeves around the power lines just before the power and data lines become close and parallel, and also again just after they separate. \$\endgroup\$ Aug 1, 2022 at 18:48
  • 1
    \$\begingroup\$ How well-shielded is the coax? If it has a particularly good shield, it would have less interference than if you had a particularly poor shield. \$\endgroup\$
    – Hearth
    Aug 1, 2022 at 18:51
  • 1
    \$\begingroup\$ The elephant in the room: Mixing signaling cables and power cables in the same conduit may be severely frowned-upon by your local electrical codes. The requirement to keep them separate is there in order to minimize the possibility of your current-limited circuits becoming suddenly non-current limited (due to a short/insulation failure), and then pose a fire risk. Check your regs. \$\endgroup\$ Aug 1, 2022 at 21:25
  • 1
    \$\begingroup\$ I guess that “signal degradation” will be more severe if the AC current has a HF noise (as an El Cheapo Power supply - from unknown brand at 1$ stores). The larger the HF current, and longer path, the noisier pickup. Using ferrite toroids at each extremity of merged AC+Ethernet could avoid major problems. How to measure and compare something sporadic as Power-On + Data transfer = assume to reach a critical threshold? I believe nobody has such answer. A final point: check for perfect insulation and visual integrity for all wire legs, to avoid galvanic AC paths= catastrophic for Ethernet. \$\endgroup\$
    – EJE
    Aug 2, 2022 at 1:07
  • 1
    \$\begingroup\$ @rslemos PS from TV should be well designed. This is not my concern. Sometimes you have a videogame close to the TV, with a Charger of Cordless joysticks: That may be noise-dirty. Sometimes a guy uses a 12V PS to light-up an LED strip: another noise-source, eventually connect to the same side of the TV rack. \$\endgroup\$
    – EJE
    Aug 2, 2022 at 17:23

2 Answers 2

2
\$\begingroup\$

It's probably fine. Digital signals aren't terribly sensitive to interference. Combined foil/braid shielded cable (RG6) would be better than just braid (RG59). Ferrite sleeves, as suggested by Math Keeps Me Busy may also help, but I wouldn't bother unless I saw a problem.

\$\endgroup\$
3
  • \$\begingroup\$ As I understand, ferrite sleeves around mains would only protect the TV set (or maybe other appliances upstream) from harmful HF interference. Or not? \$\endgroup\$
    – rslemos
    Aug 1, 2022 at 20:11
  • 1
    \$\begingroup\$ @rslemos In most cases, interference is "reciprocal". Blocking interference from device A to B equally blocks interference from device B to A. \$\endgroup\$
    – John Doty
    Aug 1, 2022 at 20:15
  • 1
    \$\begingroup\$ @rslemos Electrical devices connected to mains power can introduce electrical noise on the mains line. That noise could then get coupled to your data lines. The ferrite sleeves will help keep the noise level on your mains line down, which in turn may lessen the noise coupled to your data lines. \$\endgroup\$ Aug 1, 2022 at 22:16
1
\$\begingroup\$

As long as the shielding is good there should be any strong disturbance. Please note that shielding requires more than buying a good shielded coax cable. It is also about properly connecting the shield on both ends to the GND planes (and chassis if chassis is connected to GND). You can use either 360° bond or cable clamp but never use the pigtail for instance. About ferrite beads across the power line, that will avoid common mode current at high frequencies to flow through the power line down to the AC/DC SMPS. You can place one that will enclose both AC power cables.

\$\endgroup\$
2
  • \$\begingroup\$ All connections to be done with BNC connectors. Does it qualify as 360º bond? \$\endgroup\$
    – rslemos
    Aug 1, 2022 at 20:55
  • \$\begingroup\$ BNC connector shoud be OK since the shield is connected to ground. But it does not really qualify as 360° bond since 360° bond is realized just before entering the chassis, using a metal ring screwed on the chassis. But connection with BNC cable to a PCB should also work fine. \$\endgroup\$ Aug 2, 2022 at 21:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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