24

Let's say I have a long (> 30m) LAN cable, that connect two (very close) devices.

Could some of the unused wire, disposed as a huge coil, cause any issue ? I'm thinking about the special wire layout that would create a magnetic field (like in an inductance) and cause problems for transmitting the signal.

Large Cable Coil

3
  • I doubt it since the wire is typically insulated. If it does it would be of such an insignificance that it shouldn't be a worry. Sep 16, 2012 at 22:33
  • It looks like UTP, which is Unshielded Twisted Pair. - i.e. not insulated.
    – MSalters
    Sep 17, 2012 at 14:21
  • 6
    @MSalters It's still insulated, just not shielded
    – Dan
    Sep 17, 2012 at 14:28

7 Answers 7

40

I have done extensive testing at work with various cable and the answer is... possibly, but unlikely.

It really depends how tight the coil is - with Cat 6 - at the bend radius that it comes sold as in a box - no problems... but if you do any sort of tight bend, the chances are high that you could cause problems, but, with cat5e, you could tie it in knots and I doubt there would be any difference.

... Tested Cat 5e, 100 meters, 85 ish meters wrapped in a box, 15 meters up to this punishment:

enter image description here

and I saw no difference when compared to a shorter point to point cable.

12
  • 29
    jesus, Will. shudder what did the poor cable do to deserve that?
    – tombull89
    Sep 17, 2012 at 9:24
  • 3
    good. real life experiments are the best way to confirm or invalid myths. Something i was thinking about after asking the question : could the twisted wire helps to reduce the problem of coiled cable ? (or is twisted pair only good against EMI ?)
    – tigrou
    Sep 17, 2012 at 10:36
  • 21
    I've seen a lot of disturbing things on the internet, but this image keeps sending me shivers down my spine. Sep 17, 2012 at 12:11
  • 3
    hehe! It is really as Tigrou said, to try to dispel myths - We were testing to see how far past 100 meters we could go along with trying to break other ISO/TIA standards such as bend radius etc. .... Turns out, cat5e is pretty much rock solid... we could even put a staple right through - take it out, rub it slightly and it was back to 100%! Pretty amazing stuff really. Sep 17, 2012 at 14:11
  • 1
    There are a number of other factors to bear in mind here: Tight bends like that stress the cable, especially solid-core cable, so if it gets flexed in the future it's more likely to break (this is why there's a specified bend radius). Coiled cables are also pretty decent antennas - if your environment is RF-Unfriendly (fluorescent ballasts, transmitters, etc.) you may see problems sooner. Generally though you can be pretty abusive to a cable before it fails as Will demonstrated (you sick freak!)
    – voretaq7
    Sep 17, 2012 at 15:19
10

Theoretically, yes. However I have never seen excessive coiled wire cause a problem. I dont think ethernet really puts out enough power to create a strong enough magnetic field to cause issues. Unshielded power cables have been known to cause problems.

3
  • I've seen coiled mains cables cause a problem. Couldn't get a signal past it to save its life... Sep 17, 2012 at 0:03
  • "Coiled mains" - "signal". There's your problem. Mains cables are designed for power, not signal. UTP/STP is the other way around. The clue is in the TP, twisted pair. That's an anti-interference measure. By twisting wire pairs, any interfering signal works equally on both pairs, while the signal is encoded in the difference between those pairs.
    – MSalters
    Sep 17, 2012 at 14:25
  • @MSalters I think Ignacio was referring to data cable running past the mains coil suffering from interference (?) which I can tell you empirically can be a problem :)
    – voretaq7
    Sep 17, 2012 at 15:22
2

Rolling up excess wire in a coil can greatly impact its electrical properties and cause interference by literally turning them into inductors and antennas. but you will not effect any significant loss in signal quality by putting away an ethernet cable like this., unless you are subjecting it to enough abuse for mechanical stress to become an issue. A common rule of thumb is to never bend it with a radius under four times the cable's diameter. The coil in your image does not look like it is anywhere near violating that.

When a single wire is rolled up, magnetic fields from whatever source induce currents, adding noise to the signal. This phenomenon also occurs in ethernet cables, but the latter has multiple wires inside, carrying currents in opposite directions. Pairs of wires are generally twisted so that 'on average', both are occupying the same physical space and are exposed to the same magnetic effects. The resulting forces cancel, very similar to someone pulling equally hard on both ends of a rope on a pulley.

Rolling up LAN cables like this is perfectly safe.

0

While in theory there can be a (small) problem, in practice I've never seen it, and others haven't reported a problem either. The fact that the cable is balanced twisted pair eliminates most inductive effects in the coil.

There is a limit, though, to the length of an individual run from hub to device -- 100 meters. Over that and timings get screwed up, even if the electrical signal is solid. So if someone were to attempt this with a much longer spool of wire there could be a problem.

0

Ok, I have tested this in the worst case scenario. I work for a company that tests and refurbs cable modems, gateways and routers. Our testers sre designed to test 12 units at a time. The cable were coiled up tightly out of the way. This caused units to show a false failure on lost packets. We used shorter ethernet cables to avoid coiling and the problem has been resolved. The longer cables were re-used elsewhere without issues.

0

A coil of wire, usually around a ferrite (iron) core is an electrical choke or high pass filter. I had exposure to these building crossovers for speakers. I also experienced a coil of 300 ohm antenna wire with no core effectively filtering out UHF TV channels. When the coil was removed, the channels returned.

This simple choke coil principal backs up Janes real world tests, a tight coil of wire can filter out high frequencies and drop packets. That is exactly what we send along Ethernet cables.

If you need to take up slack, don't coil Ethernet cables, layer back and forth.

0

Environment is the critical factor wherever noise/contamination/interference is suspect. Are there halogen lights or unshielded motors/heating elements etc in the vicinity?

Are you using a DSL connection that is susceptible to external interference?

Packet loss and errors are most often caused by the environment, not just by faulty hardware. A coiled wire is basically a magnet for problems (pardon the pun).

Also, Is the UPS close by? not good, Is there an unshielded radio or speaker? Not good. Older 2.4ghz cordless phones? Not good. Flickering street light? Not Good.

To find if your environment is a potential influence for your coiled wire or DSL setup. Take an old AM Radio, tune it to the high end around (1600) and move around your area. If you hear feedback, that is RF and a cause for concern.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .