Author Options:

How to identify breaker for a currently dead circuit? Answered

There is a circuit in my house which has become unpowered, but I can't find a breaker for it.  None of the breakers in the service panel are tripped (yes, I know what that looks like), and the two labelled "not in use" don't affect this unpowered line (I've cycled them both a couple of times). 

The circuit has two GFCI outlets on it; both are unpowered and the TEST buttons have no effect (I've tried RESET, of course).  The circuit also has an overhead fluorescent panel, and an outside security light, neither of which are working.

I thought I could use a circuit tracer (a.k.a. breaker finder) for this, but what I've read so far indicates that they only work on active lines (i.e., so you know which breaker to turn off).

Does anyone have a non-obvious suggestion for how I can find where this circuit is turned off?  Or could I use a circuit tracer for this, even though it's unpowered?


UPDATE: I brought in an electrician to do the troubleshooting for me, and the problem turned out to be much easier to solve.

Besides all the kitchen lights which were out, there was also a GFCI on the other side of a shared wall which was dead. That latter one was "in series" ("upstream" in electrician's parlance) with the rest of the circuit, and it had failed, killing power to everything else.

With that dead GFCI replaced, everything now works normally. And I've learned a useful practical lesson in troubleshooting!

. When I was an Electrician/Instrument Tech, we used a unit that injected a high frequency signal onto the line (powered or not) which was picked up by a handheld receiver. Eg, this Amprobe unit. You should be able to rent a unit at a local tool rental place or an electrical supply house.


Cheap versions are fairly widely available, but they require that the circuit have power. The pro units are able to inject a signal whether the wire is powered or not, so they're more useful for finding breaks.

(I had to use one early this year to find the place where a contractor accidentally cut a cable. They did make good once that was pointed out, but I shouldn't have had to find the problem for them. Especially after they had brought a pro in to look for it. Sigh.)

Yep. I had to go to the hardware store anyway for a new floodlight. I looked at their $30 DIY circuit finder -- the transmitter that plugs into the wall uses the 120V for itself, to generate the signal on the line. As you say, the more expensive versions (like the one Nacho recommended) have battery-powered transmitters.

Hey, on a slightly off-topic note -- exactly how did you use a circuit finder to get the location of a break? When I try to design something like that, I'm coming up with either using coax and a scope to send pulses down the line and look for reflections, or some sort of differential resistance measurement. I know that can't be right. Is this worth a separate Question (and a free BA for you)?

. Timing the reflection is commonly used in industrial settings where the wiring is likely to be in conduit and/or a very long run and/or 30' up in a pipe rack.
. orksecurity's directional receiver method is more common in a residential setting.
. Measuring the resistance/impedance sounds to me like it would require near perfect wiring, but I'm no engineer.

The transmitter essentially uses the wire being tested as a low-level broadcast antenna. The detector is a directional pickup antenna and a receiver tuned to respond to that signal. Inverse-square law means that in most cases you can find some combination of location and amplifier level which lets you pick out which wire is involved and where it's going. The frequency used also tends to get filtered out by the circuit breaker's coils, so there isn't any significant amount of leakage into other circuits in the house.

To find the break, I traced the wire back from the light that wasn't working through the attic to where it went down to the basement, picked it up there (I was unduly lucky; it happened to be a relatively unusual color, which made that a bit easier), and then followed it back until I stopped getting signal. (Well, actually I first confirmed that the signal wasn't reaching the circuit breaker, so I knew the problem was definitely between breaker and switch.)

. Oh, yeah. Link is just an example. Not really a recommendation for Grainger or Amprobe (and I'm guessing you want to spend a lot less than $400 heehee), but I am a satisfied customer of both businesses.

We use Grainger in my lab for most of our small equipment purchases (like McMaster-Carr for mechanical components and materials). I'm not willing to spend that much for a single use (nor do I really want to pay an electrician for something that I should be able to handle myself).

Oh, cool! Thank you. I'll check at HD and see whether they have an unenergized-capable unit for rent.

I'll bet that one, and similar, showed up in my Google search, but I didn't even bother to read the ones that cost hundreds of dollars. The cheaper DIY ones (~$50) talked only about plugging into energized lines.

If none of the breakers in the breaker panel are affecting the outlet, it must be a broken connection somewhere.
But is it possible these are going thru a light switch somewhere?

Yes and no. In the kitchen, the overhead light box and a GFCI outlet are connected to a wall switch on this circuit. In the garage (on a common wall with the kitchen, another GFCI outlet is not connected to any switch. Outside the garage side door, the wired-in security lights are connected to a wall switch.

I believe these are all on the same circuit because they all stopped working around the same time (within a few days I had need to use them all). At the very least, the garage GFCI used to work fine, and now it doesn't reset, nor does the TEST button cause a trip (the sure sign that it's unpowered).


1 year ago

I have a dead GFCI outlet. I was advised to turn off the main power of the whole house before replacing it. After putting in the new one it should then be easier to identify the circuit.

look for something SIMPLE first. I would assume that the GFCI is failed or has moisture in/on it causing it to trip. That would cut voltage off of the line. Turn off the breaker feeding that line. LOOSEN the gfci out of the outlet far enough to do tests. Turn the breaker back on and see if you have power going to the input side of the GFCI. If you wire a GFCI, be aware that you must connect the INCOMING power to proper terminals and OUTPUT wires to the other ones. It will not work properly if connected backwards. Don't overlook the ON/OFF wall switch may be defective.

Good call! I brought in an electrician, and he found within five minutes that one GFCI, which was upstream of the rest of the circuit, had a failed hot. Replacing it with a new unit brought everything back to life. No chewed Romex, no need for a Tyvek suit :-)

CFGI should not be used in series. One CFGI ahead of the other outlets will do the job for both. Two or more in series will be unreliable and trip more often.

Are you using "series" in the same sense as I tried to use "upstream" and "downstream"? What you say makes sense to me.

I believe so. From the panel, one CFGI can have another outlet connected to it and be protected as well. This is generally how a kitchen is wired - one CFGI and perhaps several standard outlets that come after the CFGI.

I agree, and thought of those things. However, there's a pattern of failure with two different GFCIs, the overhead lights on a switch, and nearby outside lights on a separate switch. I'm more willing to believe a single wire break (or chew-through) between the service panel and a junction box, than I am to look for multiple simultaneous failures.

Note, you should not wire two CFGI in series, as you will have problems. They are not designed to be in series, and will trip.

Hear, hear! This is the existing (remodel) wiring in the house we bought just over a year ago. I have a hard time (see other comments) using the terms "series" and "parallel" properly with AC house wiring.

Apparently the fault turned out to be with a GFCI outlet on the same circuit, which failed in such a way that it disabled power to everything downstream from it. When the electrician removed that outlet and connected the two hots to each other and the two neutrals to each other, the downstream line all came back to normal.

If you were to use a battery powered device and the power was still on, you might have a problem.

Other devices designed for cable tracing include telephone line and Cat5 type cable testers (from computer supply store - around $30), and there's also a tone generator for tracking telephone/ network/ cable lines, etc.

Maybe you could feed the signal through a big diode or at least a fuse, just in case the power is on or comes on when you flip the breaker. Would be best to make your own plug to fit in the receptacle, as you would not want to send the signal through the common wire (as it would connect to all breakers).

Maybe you should find out what the different 'breakers are connected to (anyway)?
Could any wire have been chewed by e.g. rodents?


"Maybe you should find out what the different 'breakers are connected to (anyway)?"

That's why I'm confused. The previous homeowner did a terrific job of labelling the entire service panel. Every breaker is identified, and those identifications are all correct (tested when we moved in).

The problem I have now is that none of the labels relate to this particular set of outlets and switches, and neither to either of the labelled "Not in Use" breakers.

"Could any wire have been chewed by e.g. rodents?"

Crap. I sure hope not. Finding out is going to mean an adventure into our three attic crawlspaces. Time to dig out the old Tyvek suit :-(

The rodent explanation wokrs. If it's wired into any of the breakers they should either affect something also or they're not the problem?


No, whoever did the kitchen remodel (before we bought the house) did the electrical intelligently.

There are a pair of outlets inside the kitchen for some undercounter lights: one is a true GFCI, the other has the sticker saying "GFCI protected" in series with it. The other GFCI is on the other side of the same wall, in the garage, and is in parallel.

They also put the switch for the extra overhead light in the same box as the switch to the disposal. But those switches are on completely different circuits, as they should be.

Many thanks to everyone who weighed in on this. I've given Nacho the Best Answer because he gave the most direct response to the question I asked.

I've flagged Keydogstony's answer with the "sticky flag," because he addressed my underlying problem (but I don't relish climbing around the crawlspace in a Tyvek suit).

US cables have two lines, yes?

With the mains off, maybe connect a 9V battery to the end of the cable that you can identify (one electrode to each line), then prod around the potential far-ends of the cable with an LED to see if you can make it light up (maybe use two LEDs with the opposite-polarity legs twisted together, to save a false-negative due to accidentally-reversed current).

Well, there's the underlying problem. I can plug a tester dohickey (either your 9V battery, or a proper circuit-trace transmitter) into one of the deenergized outlets. The problem is that I don't know where the other end is -- that is, I don't know which breaker in the mains service panel belongs to this circuit.

What I meant was to poke at the end of every single wire in there until you got a light. There can't be that many, can there?

Oh, wait, US properties have a breaker per socket, don't they? No ring-mains?

(If the problem is inside the main fuse-box, UK regulations require the attentions of a properly-certified electrician, either to do the work or to certify the work you did as safe to connect to the mains.

Bloody safety elves...)

Ah, I see. In the U.S. the main service panel does not have any exposed contacts (violation of NEC).

As I'm NOT a trained electrician, and 100A at 120V is not something to mess with, I don't intend to remove the service panel with power energized.

With continued reflection, and very helpful input from all y'all :-), I'm becoming convinced that I have an open wire somewhere up in the attic. Sigh...