Power & Comms wiring together

[rloebrich]

There are outside metal halide light fixtures that I would like to control by an ethernet connection. If the voltage rating of the ethernet cable is equal to that of the power conductors(in this case 600 volts) can these cables be run in the same raceway. If so, where is this stated in the 2005 NEC ?

[tallgirl]

I think you're looking for 300.3 (C).

[don_resqcapt19]

Julie,
I think you're looking for 300.3 (C).
And to the FPN directing you to Article 725.
Don

[tom baker]

The key to understanding why you can't install 600 volt ethernet with power conductors is section 90.3
"Articles 5-7 can modify the general requirements in Arts 1-4"
While 300.3(C)(1) says you can. 725.54 says you can't. Everyone (even Tall Girl) seems to miss the FPN as well.
And there is the issue that the power may well introduce noise into the ethernet.

[tallgirl]

The key to understanding why you can't install 600 volt ethernet with power conductors is section 90.3
"Articles 5-7 can modify the general requirements in Arts 1-4"
While 300.3(C)(1) says you can. 725.54 says you can't. Everyone (even Tall Girl) seems to miss the FPN as well.
And there is the issue that the power may well introduce noise into the ethernet.

Tom, I didn't miss it. He has to start at 300.3 and from there decide how he wants to proceed. 725.54 states the requirements for doing what he wants to do and tells him to read 725.55. My reading of 725.55 (J) is that he could install them in the same raceway if they were sleeved. The added cost of sleeving the conductors is something he has to weigh against the cost of running another raceway out to the lights.

And on the subject of mixing power and comms, as I've said countless times, I run giant fat bundles of gigabit (and slower ...) ethernet right on top of power cabling feeding racks, and racks aren't known for being noise free. Between the noise from the power supplies, and the noise from the hardware itself, it's a very noisy environment. UTP is self-correcting -- a 60Hz induced signal on top of a piece of Cat5E running at 400MHz or so doesn't even look like noise, and frankly, if it does take a hit every now and again, TCP is a self-correcting protocol, so who cares. UTP is just very, very clean. Here are the stats from the high speed NIC on my server at the house --

#junk-box-> uptime
23:07:45 up 57 days, 20:19, 8 users, load average: 0.04, 0.01, 0.00
#junk-box-> ifconfig eth0
eth0 Link encap:Ethernet HWaddr 00:13:D4:7A:CF:9B
inet addr:192.168.0.1 Bcast:192.168.0.255 Mask:255.255.255.0
inet6 addr: fe80::213:d4ff:fe7a:cf9b/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:99958627 errors:0 dropped:0 overruns:0 frame:0
TX packets:92134286 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:79454633812 (75773.8 Mb) TX bytes:40219404613 (38356.2 Mb)
Interrupt:169

That's 200 million packets and over 110GB of data and 0 errors.

[mdshunk]

That's 200 million packets and over 110GB of data and 0 errors.Showoff. :grin:

[iwire]

Julie the I don't believe for a moment the NEC cares a bit about the error rate.

I believe they do not want a fault on the power wiring causing line voltage to be carried on the low voltage cabling.

I have seen where a factory made unit had both LAN and power conductors in the same chase. The power conductors faulted and melted into the LAN wiring. This in turn fried some of the equipment connected to the LAN. It could obviously create a shock or fire hazard.

he could install them in the same raceway if they were sleeved.

A true DIY method.

Sure would like to see that accomplished within the rules of the NEC. ;)

rloebrich, why bring the Ethernet to the fixtures?

The common method would be to put the automated controls at the power source.

[Pierre C Belarge]

To add to what Bob just posted, the NEC is not a design manual, but a safety document. If one chooses to install the power parallel with the data, that is a design choice, as long as the code requirements are met. We all have seen what unpermitted and non-inspected installations of data wiring look like....

[tallgirl]

A true DIY method.

Sure would like to see that accomplished within the rules of the NEC. ;)

Okay, what does 725.55 (J) say to you?

Here's the language --

J) Other Applications For other applications, conductors
of Class 2 and Class 3 circuits shall be separated by at least
50 mm (2 in.) from conductors of any electric light, power,
Class 1 non–power-limited fire alarm or medium power
network-powered broadband communications circuits unless
one of the following conditions is met:
(1) Either (a) all of the electric light, power, Class 1,
non–power-limited fire alarm and medium power net-
work-powered broadband communications circuit conductors
or (b) all of the Class 2 and Class 3 circuit
conductors are in a raceway or in metal-sheathed, metalclad,
non–metallic-sheathed, or Type UF cables.
(2) All of the electric light, power, Class 1 non–powerlimited
fire alarm, and medium power network-powered
broadband communications circuit conductors are permanently
separated from all of the Class 2 and Class 3
circuit conductors by a continuous and firmly fixed
nonconductor, such as porcelain tubes or flexible tubing,
in addition to the insulation on the conductors.

[iwire]

It says what it says.

Now what type of 'sleeve' do you have in mind that can be installed in a conduit with power conductors?

The typical application of 725.55(J) is inside enclosures or say surface metal raceways / non-metallic surface raceways with barriers installed.

[petersonra]

It says what it says.

Now what type of 'sleeve' do you have in mind that can be installed in a conduit with power conductors?

The typical application of 725.55(J) is inside enclosures or say surface metal raceways / non-metallic surface raceways with barriers installed.

How about a garden hose?

[iwire]

How about a garden hose?

Great if you don't feel like passing inspection.


358.2 Definition.
Electrical Metallic Tubing (EMT). An unthreaded thinwall raceway of circular cross section designed for the physical protection and routing of conductors and cables.....


344.2 Definition.
Rigid Metal Conduit (RMC). A threadable raceway of circular cross section designed for the physical protection and routing of conductors and cables...

Rigid Nonmetallic Conduit (RNC). A nonmetallic raceway of circular cross section, with integral or associated couplings, connectors, and fittings for the installation of electrical conductors.


If we ignore that there still would be raceway fill issues.

[petersonra]

I was thinking along the lines of using it to sleeve the Cat5 cable.

How would that be a violation?

[tom baker]

It says what it says.

Now what type of 'sleeve' do you have in mind that can be installed in a conduit with power conductors?

The typical application of 725.55(J) is inside enclosures or say surface metal raceways / non-metallic surface raceways with barriers installed.

I've seen this done inside luminare poles, installing coax inside LTNM conduit.

[tom baker]

TallGirl: The BISCI 568 standards call for a telephone or data cable to cross power at right angles or be 24 inches away.

[petersonra]

TallGirl: The BISCI 568 standards call for a telephone or data cable to cross power at right angles or be 24 inches away.

And this is based on what?

It is very commonplace for Ethernet and other communications cables to be well within 24 inches of power conductors and no ill effects. RS232 will not like it, but most modern comm links do not really care.

[LarryFine]

The BISCI 568 standards call for a telephone or data cable to cross power at right angles or be 24 inches away.My computer case is not large enough to comply with that.

[tallgirl]

My computer case is not large enough to comply with that.

Mine is, if I put the NICs at the bottom of the case.

I'll have to put up a photo of my server on my website. It's just a bit larger than your basic ATX case. It was selected using the following criteria --

"Hey, what's the biggest case you've got?"

Here's a pic of someone else's --

http://www.cluboverclocker.com/reviews/cases/coolermaster/cmstacker/CMStacker(1).jpg

Oh -- and my case is full, so if someone makes a larger case, that'll be the next one I buy.

[dbuckley]

A thought: Cat5 is a multicore cable with an overall jacket, and in many juristictions that passes for what you are calling a "sleeving" over insulated conductors.

[tallgirl]

A thought: Cat5 is a multicore cable with an overall jacket, and in many juristictions that passes for what you are calling a "sleeving" over insulated conductors.

I wouldn't advise putting Cat5 in a situation where it can be pulled on by anything else anywhere near it. I've never broken a piece of UTP by pulling on it too hard, but rumour has it, it can be done.

[don_resqcapt19]

Julie,
I've never broken a piece of UTP by pulling on it too hard, but rumour has it, it can be done.
You just don't pull hard enough...it breaks a lot easier than a #12 does.
Don

[sparkydon]

..of power conductors in close proximity to cat5 and coax, has anyone seen or been using the Carlon pvc nail-on two gang "dual voltage" box which contains the usual stab-in strain relief provisions on one device side, and a completely open side, with a divider plate between the two?

It seems as though the box completely seperates the ganged devices, it encourages data and power to be ran together in the same bay..

[petersonra]

you can buy cat5 cable with a metal sheath on it. i have a sample on my desk.

[tom baker]

Maximum recommended pulling tension for Cat 5 is 25 lbs.