Electrical wiring: the "last inch"

[David Nebenzahl]

The sub-thread up yonder about using 15-amp switches on 20-amp circuits got me to thinking. Actually, have been wondering about this for a long time, so here goes.

The thing is, we have circuits where the wiring and devices on that circuit are designed to safely carry a certain amount of current, for example a 15-amp circuit using devices rated for that amount and wired with #14 wire. All well and good.

But our discussions here inevitably leave out what I'm calling the "last inch". By that I mean such things as the wires that feed a lighting fixture, attached inside the wall or ceiling box, or the wires connecting a dimmer switch. These wires are always a *lot* smaller than the cable used to wire the circuit; often they're around 16 or 18 gauge.

This would seem to violate the integrity of the circuit, because now you have weak point. In the worst case, a short circuit at the device, you'd have a lot of current going through these smaller wires, until the breaker trips. Isn't there a greater chance of fire in that case?

So how does the NEC reconcile this apparent violation of the integrity of the circuit? How do folks like us who install and work on such wiring rationalize it? Is it simply a matter of practicality? (It's obviously not practical to use 14-gauge wire all the way up to every device on a lighting circuit.)

[Roger Shoaf]

Lets say you have a light fixture with a 100 watt bulb. Inside the fixture they use the smaller wires and something shorts out. Those wires are encased in the fixture and the box. The short circuit lowers the resistance and the amps shoot up and the breaker pops. The excess heat from the short is contained and dissipated in the metal fixture long before it causes mischief.

Now consider a different proposition. The circuit has a penny in the fuse box and the circuit is overloaded. The wires inside the walls heat up and catch the house on fire.

See the difference?

[RBM]

Typical dimmer is rated a 600 watts. The wiring attached to it, is of substantial size to handle its rating. If you get a 1000, 1500, or 2000 watt dimmer, you will find that it has larger conductors attached to it, to handle it's respective load. A switching device is not necessarily designed to carry the entire load of a circuit, unless it's going to carry the entire load of a circuit.

[EXT]

Also, just as a long wire needs to be a thicker gauge, compared to a normal length of wire, to carry a fixed amount of amperages, a very short length of wire can be rated to carry a larger amperage at a smaller gauge than normally used. This is the rational used on appliance cords, and the internal wiring in appliances. I have seen a formula somewhere that will determine the exact gauge needed for a given length at a specific amperage.

[David Nebenzahl]

On 8/29/2009 2:38 PM Roger Shoaf spake thus:

That makes sense; the idea is to confine any potential fires within boxes, where they presumably won't burn the damn house down.

Which is why I use metal boxes instead of plastic ones, and pay attention to properly clamping cables going into the box (rather than just sticking the cable through a hole in the box).

[David Nebenzahl]

On 8/29/2009 2:49 PM EXT spake thus:

Saying "a fixed amount of *amps*" would do.

That's not true. Conductors are rated at a certain current regardless of their length.

I was going to bring up the aspect of cords too, as our 20-amp circuits have cords plugged into them that are rated at far less than that, creating another potential source of fire.

[DT]

They specifically allow the smaller gauge wire within a fixture, and reconcile it by permitting the fixture manufacturer to act as the "Authority Having Jurisdiction" and testing their own fixtures. The fixtures must remain an "end device", and cannot have anything wired in series with them. The short length of smaller wire is proper for the limited current draw.

I believe the code used to allow 14 gauge "drops" in 20 amp circuits, which were runs from switches to lights with the similar logic that a light would never draw more than 15 amps.

[dpb]

David Nebenzahl wrote: ...

... Not exactly so--look at the voltage drop tables; at a given voltage the drop becomes excessive at a minimum conductor size and a larger conductor is required.

All these points have been considered--in essence, the answer is that the individual appliance/light/whatever has conductors sized specifically for the load.

A 100W bulb, for example, on the 15A 14ga circuit doesn't need 14ga because it draws only 1A (in round numbers)--the circuit wiring is required to be larger to account for the loading of all devices in simultaneous usage on the circuit.

Also, again, NEC specifically covers the wiring not the end devices; they're under other guidelines such as UL, etc., ...

In the end, there's no increased risk in common usage as long as you don't do something that is in obvious contravention to intended use--put a 300W bulb in a 25W rated fixture or 3 1000W hair driers on a 25-ft

16ga light-duty extension cord, say. Sure, one _can_ do stupid, it's presumed the Darwin rule will take care of that...

--

[Stormin Mormon]

How practial is it, to only plug in 14 gage cords into 14 gage branch circuits? Not very. What's worse, is trying to find a 100 watt outlet, to power your table lamp.

[Tony Hwang]

Hmmm, Is there anyone who does not do that when wiring? Then it'll fail inspection. Every thing is simple math and cool head. Actually life is.

[Tony Hwang]

Hmm, If only that cord has a load carrying 20 whole amps. I hope you still remember Ohm's law in hi school physics class. I left hi school in 1960.

[clare]

No, the ampacity of cables varies with length, becuase the resistance per foot causes more voltage drop on a long cord than on a short one.

[AZ Nomad]

Wrong. The current through a cable is constant no matter what the length as is the maximum current capacity of a wire.

Resistance and voltage drop are a separate matter.

[Charlie]

Seems to me there are two pertinent factors.

One is voltage drop which is an IR drop and really only affects the operation of the device at the far end.

Second is the power dissipation in the wire that is the I(squared)R value. That dissipation is spread linearly along the conductor and gives the rise in temperature that is relevant to ignition.

Charlie

[Doug Miller]

Yes, exactly so. Voltage and current are not the same. A long conductor *does* cause voltage to drop, but it does *not* affect current.

[Doug Miller]

Nonsense.

Voltage and current are not the same.

Conductor ampacity is dependent on:

- the conductor material (Cu vs Al)

- the insulation material

- the conductor diameter and NOT on the conductor length. If you think rated ampacity is in any way dependent upon conductor length, then please cite the section of either the CEC or the NEC which says so. I'm not going to hold my breath.

[bob haller]

certinally a 18 gauge light cord on a 20 amp circuit is a fire hazard.

the plugs of all such cords should be required to have a built in fuse.

I once had a customer using a 18 gauge ungrounded light extension cord on a 18 amp grounded machine that tended to burn wires off inside the unit.

I wrote it up as a safety hazard, warning them buy a AC extension

next time I went they had improved things by stapling the extension cord to a carpeted wall:( wire got very hot showed signgs of melting.....

I fixed the machine then went shopping in the mall, and bought a AC extension cord. the customer got mad when I cut their junk cord in pieces and gave them free the air conditioner cord. told them the new extension cord was far cheaper than even the paperwork for the insurance claim when they burned down south hills village.

the customer was pissed and never called me again. frankly i didnt care.

have you noticed christmas lights now have fuses in each plug? that should be required for all plugs

[JIMMIE]

It does not affect the current that the wire COULD carry or the AMPacity. A long cable, compared to a short cable, will the current in the total circuit because it adds resistance to the total circuit. Voltage E, Current I and Resistance R are not independent of each other under normal conditions.

E=3DIR I=3DE/R R=3DE/I

Jimmie

[JIMMIE]

I WROTE WHAT!!!!! Sorry answered two phone calls and three questions from my wife while trying to write this. Please ignore my dribble.

Jimmie

[Doug Miller]

I normally do anyway...