On residential new construction, how close can a interior phone line be to electrical wiring without being affected by electromagnetics? Can they cross each other?
- posted
17 years ago
On residential new construction, how close can a interior phone line be to electrical wiring without being affected by electromagnetics? Can they cross each other?
The "book" on phone wiring says you can run them perpendicular but not parallel to each other. I've run them parallel in the same drilled holes in studs for short distances with no problems
If you use twisted pair phone wire there is no limit.
Others will tell you different, but I have ran phone, cat5e and Romex all together. In fact, when I wired my garage, the phone cable (I used cat5e cable) was ziptied to the main run of Romex cables for the whole garage. No problems with it at all. My phone cable is brought to my house hanging underneath the main lines, and people want you to believe that a 120V cable is going to cause a interference problem. nada!
It would if you were using the old four wire, non twisted cat 0 phone wire. Since you are using cat 5 wire which is twisted pair the twist cancels out the common mode interference that would most certainly have been picked up with the old untwisted wire.
As for the drop to the house which is likely non twisted drop cable, it is not in close enough proximity to be a problem. The magnetic field that induces interference drops off in strength rapidly with distance.
Pete C.
I realize the original post was about phone wire, but I'd be interested to see a data analyzer hooked up to some cat 5 data cables that were run next to electrical wires. I'd be curious as to how many dropped packets and retransmissions due to errors would occur. The average user might not ever notice a problem but errors like these increase network traffic and would drive a network engineer up a wall.
Pete C. wrote:
At 100mz I say ZERO. I have not tried the 1gz LAN. I did a lot of testing with Ethernet and Token Ring cables in my old office. I was not able to reproduce any of the urban legend "interferance" with flourecent light ballasts, mixing phone and data or paralleling 60hz power lines. This was using the IBM LAN sniffer program and the normal office traffic on the LAN along with me shipping big files around.
Thye can be right next to each other abd touching without any electrical interference, but it won't meet code on safety.
Not sure about the interference, but you're right about not meeting code. The NEC, in Sec. 800.133, specifies a 2 inch separation between "communication cables" and power cables unless the cables are somehow physically separated, such as by a conduit.
Remember that if the is no current on the AC lines, there will be no magnetic induced fields, even though the lines may be energized.
Since the dawn of telephone communication, crosstalk between the phone lines (and from external sources) has always been a problem, although, now it has been minimized due to a better understanding, and improved materials and installation techniques resulting from scientific developments.
The first phone lines were single conductors with a common ground like telegraph lines and suffered from severe crosstalk and power line interference. Multiconductor circuits, loading coils, and twisting of conductors helped to solve the problems. Ultimately, phone lines and switches became digital (ESS) long before the Internet became mainstream.
Beachcomber
Read 800.133(A)(2)exception 1, the jacket of NM cable is "separation". The 2" refers to individual conductors.
I stand corrected.
Agreed. If twisted to at least Cat3, running along side a long run of romex (for example) for the ENTIRE length of even the largest residence (rare) presents no degradation of signal. Separation is always best but not always practical.
Agreed, but I think your concern is EXACTLY WHY Cat5e was developed.
I suspect that, except in isolated cases, electromagnetic interference with Cat5e is virtually non-existent. I'd bet that you'd find more drops and retries due to poor connections such as at routers, patch panels, and improperly made/crimped RJ45 ends, to name a few likely spots. Just a thought...
In my BICSI training they pointed out virtually all of the crosstalk happens in the last inch of the run, where it is made up in the keystones. This is usually because of poor termination practices.
Interesting. The twist rate is much larger than an inch, no? I wouldn't think a problem with the twist in the last inch wouldn't matter much.
It is so common they gave it a name, Near End Cross Talk., NeXT
No.
In ethernet/Cat5 cable, the twist rate is several, full twists PER inch. Someone more familiar with LAN cable probably knows the rate.
It would. ...and apparently does.
I have spoken with several techs in recent years that claim it is CRITICAL to CONTINUE the twist ALL THE WAY to as "flush with/against" the connector as possible.
The twist rate is actually a littler different for each pair in that cable.
that is true. It is best to not disturb the twist until you are ready to make the punch andf then only as much as necessary. At slower data rates ir will probably never affect you but as we start getting up around a ghz on copper it is the difference between working and not. It wasn't that long ago that people were saying 100mz would be the end of the line for twisted pair.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.