The Circuit Detective    -  Pinpoint the Problem

<<previous   HOME   Sitemap

Find the Source of Your Electrical Problem

2. (Troubleshooting Strategy) Locating the Fault

Of the causes we have just gone over, you should now know which one you are dealing with, thanks to that Previous page. To find the exact place causing your electrical problem along the circuit, you will want to go to the corresponding section of this current page:


The Short Circuit. Since the short, the ground-fault, the arc-fault, and the shock are all faults, that is, cases of unintended continuity, the procedures for pinpointing their fault-points are similar. In general, this involves isolating parts of the circuit from each other and then retesting for the continued presence of the continuity. While this can be accomplished by a divide-and-conquer approach, I will suggest some more efficient ways as well.

A hot-to-ground short (breaker-tripping ground-fault) is more common than a hot-to-neutral short. It can be helpful to know which you have, and there is a way to know. But let's consider how to attack the short in any case. Most breakers can stand up to repeated shorting, so be ready to keep resetting the one in question. Don't HOLD the breaker on, however; just quickly and firmly push it on.

First, on the dead circuit, unplug everything and turn all on/off switches off, and turn only one switch in each 3-4-way system the other way. Try to reset (push off, then on). If the breaker stays on, one of the items you disconnected from the circuit has the short in it, so reconnect one at a time, turning power back on each time. That should identify the culprit. If the short is in a string of lights, keep reading.

But if the breaker retripped in spite of all the disconnections you made, then something more permanent is shorting. I would suspect outdoor things before indoor. In any case, pick a point along the circuit -- maybe midway along -- and disconnect the hots at that point. This is "divide and conquer." If the short remains, it is electrically closer to the panel than that chosen point. If the short is gone, it is further out. By reconnecting what you undid and then opening new hots in the direction of the short and by keeping track of all this, you should reach a place or a particular length of cable at which to look for the actual short.

If you suspect a recent screw or nail is to blame, see Screw. If the short's location is inaccessible, you might know enough to bypass it with new cable. Overall, if the short circuit just won't reveal its source, a good electrician might be able find it.


The Ground-fault. [For the arc-fault see AFCI breakers.] By ground-fault here, I mean a ground-fault that has tripped a GFCI receptacle or a GFCI circuit breaker. (But realize that a GFCI breaker in the panel could be tripping for an overload or for a hot-to-neutral short instead.) Review GFIs.

You have the advantage that the number of downstream loads is limited and their locations are knowable or guessable. Namely, the dead receptacles will tend to be found at the places that were required to be GFCI-protected. You have the disadvantage or complication that the fault could be from hot to ground or from neutral to ground. To determine which it is, you can temporarily disconnect the "load" neutral(s) at the GFCI; if it still trips, the fault is hot-to-ground. Otherwise it was neutral-to-ground. Yes, a GFCI will trip for either condition. Either one provides an alternate path for some of the load current. To the GFCI, some current is missing when it compares the amount flowing on black and the amount on white. If there were no fault, these would be equal.

For either form of ground-fault, note whether any of the dead receptacles have a cord plugged in and unplug them. Then see if the GFCI will reset. If not, note whether any of the receptacles are broken and whether they, or the interiors of their boxes, are wet and whether any out in the yard receive their power by a buried cable. Replace any broken ones, dry out the wet things and reset the GFCI. If it still trips, undo the hots and neutrals of the buried cable at the box where it seems to leave the house. Reset the GFCI. If it no longer trips, you need to reconnect and then repeat this disconnecting procedure at other boxes in the yard. If you have found nothing making contact from hot or neutral to ground or earth, then you may find the GFCI still tripping for one particular piece of buried cable. So you would pull up, dig up, and repair or replace that piece.

If the GFCI tripped even when the line feeding to the yard was undone, come back indoors. Open any dead boxes. Look for hot or neutral wires making contact with any ground wires.


The Shock. CAUTION: This is one electrical problem NOT to observe yourself, at least not by using your body directly. I recommend a neon tester to check metal for hotness. For this sort of ground-fault, two strategies are possible. One is to leave the ungrounded fault in place and locate it first, grounding things better later. The other is to provide a good ground to the thing that delivered the shock. This will probably create a breaker-tripping short, and then you can deal with it as a Short. Though it feels less safe and I have to be more careful about myself, I find it more efficient to use the first strategy. You should attempt what is safe in your judgment, according to your knowledge. I once had to hunt for a shock condition which had energized all the metal-sheathed cables, all the pipes, and all the ductwork throughout a house and its basement.

If you have decided to leave the shocking thing(s) hot, first see which circuit, when turned off, eliminates the hotness. Get well acquainted with all the other things that are part of that circuit. Then turn it back on. See if those other things show that stray hotness, including the grounding hole of receptacles. If the home was built after the 1960s, hotness is more likely to be limited to one thing or to that fraction of the circuit from which a ground wire has become disconnected. If the home was built before the 1960s, it is likely to spread hotness to various metal things if metal-sheathed cable has lost its contact with ground.

If nothing else but the "shocker" shows hotness, disconnect the hot wire of that one item. If hotness disappears from its shocking metal, then it has the faulting wire or part within it. If hotness persists, then on this circuit unplug everything, and turn all on/off switches off, and turn only one switch in each 3- or4-way system the other way. Did one of these actions eliminate the hotness? If so, it is the home of the shock.

Look along this circuit for any broken receptacles, like where a too-long silver cover-screw may have broken the receptacle's plastic apart. Also, take all covers off receptacles and switches of the circuit, and look for a ground wire curled up next to the hot terminals.

If the shock -- the hotness -- is not always there, is there some automatic appliance or light that is responsible when it turns on -- at certain times of the day, for instance.

Beyond all this, pick a point midway along the circuit and undo its hot wires there. If the shock-location hotness disappears, then the fault was coming from somewhere electrically beyond (away from panel) this midpoint. If the shock-place is still hot, the fault is electrically back toward the panel from midpoint. You can reconnect the hots and then repeat this divide-and-conquer procedure at other points to narrow the fault location down. To avoid confusion, record your results as you go.

If nothing leads you to the fault itself, you can give a good ground to the shocking thing, so that a short is created and perhaps trips the breaker. Then troubleshoot it as a Short.


The Open. See definition. I estimate that the chance of an open happening in a given household during its lifetime is at least 50%. Since this is perhaps the most common electrical illness homes suffer, take heart! Hundreds of opens are solved around the country every day, and yours has no reason to be specially stubborn. Here is my electrical advice. Since an open is an unintended discontinuity, locating it can involve experimentally disturbing connections till the bad one makes good contact again briefly, or else figuring out fruitful places to look for the discontinuity.

First, you need to learn all you can about the extent of the outage and its circuit, so that you will be ready to probe into enough places and won't have to probe into any uselessly. Do not skimp on this step. Be thorough. You will be comparing the lay of the outage to the lay of the whole circuit. (You should already have narrowed your open down to just one circuit, not having the symptoms of a Main open.)

So your first job is to know the outage. How well would you get to know it if I had to pay you $100 for every dead item you found, but you had to pay me $500 for every one you missed? You are very aware of the lights and outlets you no longer have the services of. What about the ones you never use and the ones behind furniture and stored boxes that you don't even remember exist? Are you going to overlook items that are unique or out of the way: the doorbell's transformer, a wired-in smoke alarm, or junction boxes in the attic or crawl space? To be the source of your problem, it doesn't have to be something you use. Yes, you may have to get more intimate with your home than you were in the past. Document what your new intimacy reveals.

Your next job is to know the outage's whole circuit, if possible. For this you will need this same thoroughness and perseverance you exercised for knowing the outage. Yes, it is most likely that several more things of the circuit in question are still working. Don't rush to see what the labels on your panel say. They won't say enough and might even be lying. You can't afford to base your whole investigation on shaky assumptions. I don't care if you labeled it yourself, back when everything was working. You may have to do a little guessing later on, not yet.

And how can you tell which circuit the outage is part of? If it is the hot wire that is open -- not getting through from the bad connection onward -- you may not be able to learn its circuit. Yes, you could try to believe the panel, or you could turn off other circuits one at a time and see which one "seems less full" than the others. But you have a 50/50 chance that it is the neutral (white) that is open instead. If it is, you can absolutely know which circuit you are dealing with. You can -- if you get a tester. A neon tester, non-contact voltage checker, or (for circuits wired since 1970) a receptacle tester, will let you do this. They cost from $2-$20. How will they tell you whether the neutral is open? By lighting up or beeping when inserted in a non-working outlet or one of its slots. You know the neutral is bad from two facts: that the outlet can't run normal things and that it has a good hot, because the tester lit up. If the hot is good, then all you need to do to identify its circuit is turn breakers off one at a time while continuing to look for the light or beep of your tester.

If you were lucky enough to have an open neutral, don't slack off. Keep your luck going by doing as much research on the full extent of the working items as you did on the non-working ones. Notice, I don't call them dead, because in your case they have a live (hot) wire at them, which is quite able to shock you when you have the circuit on. Again record everything on the circuit that works, including those out-of-the-way items.

Good. And you folks with the open hot, don't despair. You will have your day. And you tester-less people, don't count yourselves out. Now, whether your open is of the neutral or the hot or unknown, you are going to try to disturb the bad connection back into working briefly. I call this The Jiggle Method. If you know the circuit's working things from its non-working ones, or can only guess that nearby working things may be part of the same circuit as your non-working things, you can now get down to the real work. Have all circuits turned on. Go plug a working, turned on lamp or nightlight (radio?) -- something that will work instantly without a lag -- into one of the non-working outlets. (If the outage involves no outlets, turn on the non-3-way switch of a non-fluorescent light within the outage area.) Then assign someone else the job of watching that light constantly and reporting to you immediately if it tries to light up -- even the slightest flash. I suppose you could drag a light on an extension cord around with you, all by yourself.

Ready? Position yourself among all the non-working items you discovered. Face toward your electrical panel. As you move in your imagination, without regard to walls and floors, toward the panel, the first working items (of the circuit, if you know it) that you would start to pass into the midst of are crime suspects, along with those non-working items nearest to these working ones. You, the detective, are to go take the covers off all outlets and switches -- dead or alive -- along this Dead/live border.

Now use The Jiggle Method. Video. Plug something else into these outlets, one after another, and wiggle it side to side somewhat. Next go back to the same outlets, and also to the switches, and stick a strong, thin stick of plastic or wood, not metal (e.g., a chopstick) beside the device, pressing firmly on the wires you see, and then stick it more behind the device, pressing and poking various wires you can't entirely see. You can even pound on the wall or ceiling near the boxes of this electrical border. The purpose of all these activities is to disrupt wires back into good contact. If at any time your helper tells you the light flashed or stayed on, stop where you are! This is where the connection needs to be improved!

This Jiggle Method will often succeed, but certainly not always. This is the point at which you sad people with the open hot get a shot of hope, and the open neutral folks have to sit down and rest awhile. The reason is, if you invested in one of those non-contact voltage checkers, you can now try letting it tell you where the open hot is. You see, a 50/50 chance has come around to you now. There is a 50% chance that the open is at the first non-working box along the circuit. The other 50% is that it is at the last working one. Therefore, if you were to stick that volt-stick past all non-working receptacles or switches to the wires behind them, you have that 50% chance that it will register steady hotness, telling you that you have found the bad spot! Be sure it gets up against each of the wires in the box. You will probably have to loosen the device from the box, somewhat, to do this. If no non-working device shows this hot wire behind it, then it will probably be one of the exposed working devices that hides the electrical problem. You will need to visually and manually check or improve the black-wire connections at these places, with the circuit turned off.

You open neutral people, however, will have to resign yourselves to checking and improving the white-wire connections at both the non-working and the working places on this dead/live border (shown opposite). Those who cannot be sure what things are part of the outage's circuit will have to check and improve connections of both colors in the general area of the outage and of nearby working things, with circuits off.

When I say that 50% of the time the open will be found at the first non-working item and 50% at the last working item, I exaggerate only a little. The exceptions are: junction boxes (which can't be said to "work"), rare inaccessible splices or breaks, the chewing of rodents (also rarer than imagined), and underground splices. Nevertheless, generally, the trouble point will be found in electrical boxes, including the panel.

More possibilities. It was common in homes built 1940-1970 to run a circuit through light boxes more than outlet and switch boxes, so there may be connections to be disturbed and checked there. Homes built 1900-1950 may still be relying on their original knob-and-tube wiring connections. If these were not soldered well, they can be the cause of an open. They are not found in electrical boxes but in ceiling and wall spaces. If they are accessible, someone jostling them with a stick could make the light, in The Jiggle Method above, flash on.

A restatement of how to find an open is found at the Diagnostic flowchart or in the Diagnostic tree. If you have tried these things without success, a good professional may be needed. That includes me.

To make repairs...


"I found you by way of Google, and thank heavens I did. I was able to solve 2 issues I inherited. The tutorial and trouble shooting sections were invaluable, and really helped to demystify the circuit I was working on." -Jim

Short-circuit article by Wikipedia

© 2007-2013 Larry Dimock

<<previous   HOME   Sitemap