Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Remodeling my house and want to wire up each room with Ethernet/coax wall ports. Both Ethernet, coax and power will be running through the sub-floor.

I've read that there needs to be 2 feet between power cables and Ethernet to limit/prevent interference.

Is this true? If not, does anyone know what the distance is?

share|improve this question
1  
Interesting question. I've never heard of the limit and in some installations I've seen, cables were next to each other and network worked. I didn't do any speed tests, so I don't know if there was any impact. –  AndrejaKo Mar 20 '11 at 19:24
    
@AndrejaKo I had a hard time even finding that on Google, so I'm unsure of how true that is. Hoping someone here might know. –  Adam Kragt Mar 20 '11 at 19:27
    
Found a reference! –  AndrejaKo Mar 20 '11 at 19:35
    
@AndrejaKo Thanks! It's a very informative post, even if it is from 2003. Now on to reading weather or not to do Cat5e or Cat6 :-P –  Adam Kragt Mar 20 '11 at 19:55

6 Answers 6

up vote 3 down vote accepted

It seems that there is some truth to interference, but as far as I'm able to see, there is no real information available. There are some best practice tips and similar, but it all comes down to few things like: Use highest cable rating you can afford, use STP cable and 8P8C connector whose rating is same or higher than cable's rating. Be sure to use STP connectors with STP cables, or you'll get negative effects. If you can, use cables designed for outdoor installation, since they have better shielding. Make sure you don't bend cables too much. Bend radius should be 8 times outside diameter of the cable. Make sure that single cable is not longer than 90 meters (for main cable) + 2*5 m (for patch cables connecting the equipment).

As for power line interference, I couldn't find some concrete research. It appears to be affected by the clarity of the line signal, so anything which can cause disruptions such as some dimmers and similar devices can have negative effect. I've seen recommendations for distance to be form just 15 cm to nowhere near the power cables.

On the other hand, there are reports from people who used unshielded Ethernet cables near power cables and near devices which are well known to cause interference and who had no problems.

In the end, I'd recommend that you run a single cable in conditions of normal use and do some speed testing and decide if cable is a problem or not.

Here are some references I found: TechRepublic, Cables To Go, MetaFilter, LinuxQuestions, Pirate4x4, SooperArticles.

share|improve this answer
    
Thanks! Very much appreciate the sources. –  Adam Kragt Mar 20 '11 at 20:10

When cables run in parallel, the electromagnetic induction of one cable always affects the other. That is normal. If cables cross perpendicularly, theoretically there is no interference.

You send normally a signal over two wires. If the interference affects both cables, there is an annulation effect, therefore the signal arrives fine. If only one of them is exposed to the interference, the signal suffers degradation. This effect is normally avoided by twisting the cables (check the Ethernet Cat5 pinout, the two Tx/Rx are twisted).

High-speed and low-voltage signals are of course more sensitive to interference. The interference level depends on many factors (cable quality, position, distances, type of noise, frequencies, etc.).

The structured cabling standards recommends 10cm of distance for Cat5 and common power cables running in parallel, I think (that's the function of a standard).

But as I said, anything depends on the situation. Once I run an ethernet cable on the same duct with a power cable, 60m (Ethernet standard max distance is about 100m), and still works afaik. Another time we installed Ethernet cables on a roof full of fluorescent lamps, following the 15cm distance and the crossing angles rules. We needed to decrease the speed to 10mbps to be able to receive something, and it still does not work well.

share|improve this answer

Most cable management systems built for this purpose separate the 2 by no more than about 9in. There might be some crosstalk, but if you think in most office buildings they bundle the two together in the same 6in wide trunking.

share|improve this answer

Don't let anybody kid you - your best answer above suggests a doubt about the effects of electrical interference. WOW.

I have been involved with electronics and acknowledged as expert repeatedly by others since 1972.

Electrical interference is a VERY REAL problem

Make no mistake, the handling, bending and even tying down of Ethernet cables without due care can all introduce inefficiencies and/or disruptions in the performance of any network. Even if your network seems to be working ok - is it really? I assure you - if you violate certain considerations - your network performance WILL SUFFER.

My experience suggests that electrical interference introduced by a wide array of potential culprits is by far the most deleterious bane of many network messes I have had to sort out over the years.

My rule of thumb is to buy yourself a good one ft. separation between any AC line and any parallel run of unshielded Ethernet cable. In moderately noisy locations - you may get away with 6 inches - I would never run any closer. You should also avoid crossing an AC cable at a right angle within a few inches if, possible.

I have seen people do such stupid things as resting transformers directly on patch cables... and then swearing off their ISP or computer tech.

Stop and think what you are asking the technology to do - or do you not know - and you'll quickly understand why it is in your best interest to do everything possible to avoid causing issues if you want to get the performance you're paying for with your equipment investments. Not paying respects to the care of network cabling and its installation is anathema to plugging your system unit into a surge suppression or UPS device that isn't grounded on the other side - you're just shooting yourself in the foot while pretending you haven't.

Never buy the argument that someone else did something and it didn't affect them. It's almost always provably wrong. Let me recount the number of people who failed to heed advice regarding the electrical protection of their systems - and the actual ground quality of their outlets.... oh - I don't have that many spare years for the typing.

If you are installing this stuff yourself I would STRONGLY advise tou to further research and ensure you have the specifications for terminating Ethernet cables properly AND FOLLOW THEM TO A 'T'! If you doubt at all, the quality of termination you have just made - redo it - it isn't worth later frustration. Make sure you do not sharply bend your cables at any point (the published standard is a minimum bend radius of 4x the diameter of the cable for UTP and 8x for STP.) also - do not pull with excessive force or jerk the cable to get it through or by some obstruction while pulling.

The only physical difference between 10MbS (older cable) and Cat5e is the amount of twist on each pair and the pairs among themselves within the jacket of the cable - THEREFORE - you are kidding yourself if you think distorting the shape of the cable by insensitivity during install can't come back to bite you later.

Pay attention to detail and you will find yourself well pleased in the end. Otherwise you may still get things to work - but you may not be getting the speed you think you are. I should note - there is no Broadband service in the states yet that will challenge even a 100Mbs system - but you'll pay if you have to do many large file transfers between units on the network.

share|improve this answer

I have never heard of two feet as being the standard so I can not confirm that. However you are correct that Electromagnetic interference can cause problems with unshielded twisted pair cable. It is definitely best practice to keep power cables as far away from Ethernet cables as possible, so I guess two feet would be about right. It shouldn't be able to outright stop data transmissions, but it can cause a variety of speed and data corruption issues which can be quite frustrating to diagnose later on. If distance is an issue you can use Shielded Twisted pair (STP) cable that is protected from outside EMI, it's a little more expensive but overall the cable quality is much higher.

Obviously it is best practice to keep the power and data separated however in real world tests EMI will not grind your network to a halt. You may see random hiccups with network connection or while streaming movies but nothing too serious. So there is no need to be obsessive about keeping them two feet apart. Below is the results of a test done on the affects of EMI on UTP by Siemons:

Packet flow on the Fast Ethernet hub was monitored via NetXRay Protocol Analyzer and Network Monitor software7. The software not only monitors network utilization levels and packet errors, but is also capable of generating errors on the network.

The operating systems under evaluation were charged with the task of simultaneous video and file transfer to keep utilization rates at approximately 30%. The NetXRay software confirmed utilization rates. Under normal operating conditions (no known sources of EMI), no packet errors were detected for any cable type.

To demonstrate the effects of packet errors, the NetXRay software was programmed to generate CRC, fragment and alignment errors. Depending upon the size of the defective packet, resultant errors ranged from a slight pause or glitch in the video display to a complete network ‘crash'. Typically, CRC and alignment errors, associated with the video transmission protocol, resulted in the most serious network crashes. All network errors were visually detectable on the video display.

So in a test environment with artificially generated defective packets the possible affects of EMI could be seen. However the conclusion states they were unable to cause issues using actual EMI:

CONCLUSIONS

No packet errors were detected for either the ‘generic' or ‘enhanced' category 5 channel configuration regardless of EMI source type, source location, or duration of exposure.

Here is the article for further reading.

share|improve this answer
    
Thanks for the article! –  Adam Kragt Mar 20 '11 at 20:11

Yes.

Data consists of electrical impulses on the wire, so a strong electrical signal nearby can degrade or interfere with data transmission. Ethernet is very susceptible to EMI (electromagnetic interference).

If you must run your electrical and your Ethernet near each other, you'll want to use a higher-grade type of cabling called STP (shielded twisted pair), which is more expensive.

To enjoy high data throughput in your network, keep your power as far away from your data cabling as possible. Cat5/Cat6 cable should not be run parallel to electrical lines, but may cross them perpendicularly.

share|improve this answer
    
Thanks! I'll keep that in mind. –  Adam Kragt Mar 20 '11 at 20:12

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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