DIY Home Improvement, Remodeling & Repair Forum

DIY Home Improvement, Remodeling & Repair Forum (http://www.houserepairtalk.com/forum.php)
-   Electrical and Wiring (http://www.houserepairtalk.com/f9/)
-   -   Preventing noise in Ethernet cables (http://www.houserepairtalk.com/f9/preventing-noise-ethernet-cables-15346/)

drewdin 01-10-2013 03:03 PM

Preventing noise in Ethernet cables
 
Hi All,

In multiple places the electrician ran the ethernet cable along side the power cables in the room, I want to try and prevent as much noise as I can, I have read that all I need to do is cover the areas with tinfoil to prevent noise.

My question is, is it that easy? Will tinfoil work or is there something else I can do that will work better? Removing the cables and re-routing them is out of the question.

Any recommendations will be greatly appreciated, Thanks!

JoeD 01-10-2013 06:05 PM

Around the Ethernet cable shouldn't hurt but do not wrap the electric cables with tin foil.

drewdin 01-10-2013 06:43 PM

I only planned on covering the ethernet cables, just to verify. Is that the best way to prevent noise? Is there anything better that you would recommend? Thanks

nealtw 01-10-2013 07:24 PM

If that work's, would that aluminum tape they use for sealing ducts work, it is sticky allready. No not duct tape!!

Wuzzat? 01-10-2013 07:43 PM

I searched on

"bit error rate" romex ethernet

but didn't come up with a document that related the three in a way useful for you.

I'd recommend cutting wood on a table saw powered by an extension cord or a length of Romex that your Ethernet cable is wrapped around for a few feet. If your data transfer rate doesn't slow noticeably because the parity bits don't match then I'd say you're OK with the installation.

Supposedly there is also smart home cabling that runs both power lines and data lines together.

Non-ferrous metal (foil) will not shield against the primarily magnetic field (H field) radiated by power cables. Foil or braid attenuates or reflects electric fields (E field), like nearby radio station transmissions.

CallMeVilla 01-14-2013 12:25 PM

Shielded Cat6 cable is the minimum quality you should use. It is designed to protect itself from outside interference, BUT, it is recommended to NEVER run them side-by-side to your electrical wiring (Romex). Typical is to run electrical down one stud and the Cat6 down a different stud.

If you ever have to cross the Romex, make sure you do it at right angles (perpendicular) so you minimize the chance of interference. I ran 4,000 ft of CAT6 in a smart house and used this approach throughout. No interference issues developed!

Good luck! :D

drewdin 01-14-2013 12:48 PM

@vanilla, the cat5e cable has already been run. I cant change its location now, all i can do is try to prevent any noise. Any suggestions?

Wuzzat? 01-14-2013 12:54 PM

Quote:

Originally Posted by CallMeVilla (Post 82210)
BUT, it is recommended to NEVER run them side-by-side to your electrical wiring (Romex). Typical is to run electrical down one stud and the Cat6 down a different stud.

ran 4,000 ft of CAT6 in a smart house and used this approach throughout. No interference issues developed!

Can you post links for this, preferably by the IEEE (no Mike Holt)?
The science may say it would make no difference, running parallel or not, but I don't know for sure what may happen in the real world.

In the 4000' case there may not have been interference in any case (the ad hoc fallacy) but this issue comes up frequently enough on forums for me to want some science links to verify.
You may after all be right, and this particular issue has never came up in my work so I haven't dug into it.

My main misgiving has to do with the qualitative separation of 60 Hz sine wave magnetic fields vs. the high frequency square wave electric field (which is totally contained within the shield) used in data transfer.

If there is interference, drewdin, you may be able to get a "60 Hz notch filter" to reduce it.

CallMeVilla 01-14-2013 01:06 PM

Happy to share ... Here are some links that all poride bits and pieces of useful information:

ROMEX and CAT6: http://www.tek-tips.com/viewthread.cfm?qid=1416430
HDMI and other issues: http://www.remotecentral.com/cgi-bin/mboard/rc-custom/thread.cgi?16953
Install issues: http://www.electrical-contractor.net/forums/ubbthreads.php/topics/198957/all/Cat6_installation.html

Wuzzat? 01-14-2013 04:14 PM

The plot thickens: look at claim 1 of
http://www.freepatentsonline.com/6998538.html

"We claim:

1. A hybrid electrical cable providing for power transmission or distribution and low power signal communications, the hybrid cable comprising: a) a power cable including a first group of one or more conductors for conducting power; b) a power cable insulation jacket overlying the first group of one or more power conductors, the power cable insulation jacket including a first layer comprising a soft magnetic material having a coercivity of 1 oersted or less; c) a group of one or more low power signal conductors disposed exterior of the power cable insulation jacket; and d) an outer jacket overlying the power cable, the power cable insulation jacket and the group of one or more low power signal conductors, the outer jacket comprising a metallic material."

Apparently Ulectra Corporation (North Miami Beach, FL, US), the assignee, is producing this thing.

So here's your ferrous metal for mag field shielding (use BX cable!) and the patent does mention shielding later on as a design constraint and so drewdin may have to resort to extraordinary measures to keep his Ethernet functioning to full capacity.

Or, he may luck out.

And this
"by absorbing radio frequency (RF) electromagnetic emissions generated by high voltage, high frequency electrical transients which may be present on one or more power conductors of the power cables due to external high frequency electrical disturbances."

implies that the problem is not so much the 60 Hz but that the Romex acts as an antenna to pick up, or conduct, all kinds of electrical noise.
In this case the Ethernet may work better late at night when the PoCo customers are asleep.

BTW, when I left the USPTO the patents were just beginning with the 5,100,000 numbers. IIRC, they had 20% turnover each year. . .and I lasted a few years.


All times are GMT -6. The time now is 03:19 AM.