The Circuit Detective

Stories From the Trenches

Adventures in Residential Electrical Troubleshooting

October 23, 2006

  A busy day. First and easiest, Mrs. Shaw. All she had done was unscrew an indoor flood from a recessed can. And I saw that the bulb was now hanging by some wires. Usually that means the glass broke loose from its metal threads and I need to use my needlenose (spread outward) to remove the metal end from the socket. But this was different. Never seen a bulb DESIGNED to unthread from its metal end. This was a compact fluorescent flood and it got ruined by this unscrewing. I replaced it.

  Next a sad story. This guy's mobilehome had oodles of electronic stuff. Most of it got fried back in June, he said. So the power company thought they found a bad neutral in the transformer and replaced it. Now his next set of stuff fried, some in spite of good surge protectors. Sustained higher voltage isn't a surge. It's a melted icecap, not a little thing like a tsunami. Lights were getting bright, then dim, some of them popping. He was there. This wasn't happening when I arrived. OK, sounds like the power company was wrong. But we have to be sure the bad main neutral isn't in my customer's meterbox or panel. The panel had no 120 volt breakers sharing a neutral and the damage wasn't limited to two circuits anyway. The panel neutral was tight and looked good. There is this rigid strap that connects the neutral bar on one side to the other. I thought I saw a little discoloration near one screw holding it together. Nope, it was tight and OK underneath. How about a big junction box underneath the home for the main feed to the panel? Some mobiles have this. I crawled under enough to see there was none. Then what about the outdoor meter/disconnect on his pole? Neutrals in and out of there checked out. There was even a feed from there directly to his shop building. Did anything out there get hurt? He said it did in June but he hadn't checked this time. We went out there. Garage door opener was dead. Hmmm. This confirmed that the problem was back ahead of the meter setup. I called the power company. But I knew they might not take this comes-and-goes problem seriously enough. And the guy needed to be able to use his computer (it escaped) without anxiety. So I proposed putting all his 120-volt breakers onto the same phase in the panel. That's what we did. That would keep a bad main neutral from getting at them from the other phase (see main opens). Right as I finished that, power company shows up. This lineman is used to false alarms. He considers everything I have to say. And the fact that the neighbor off the same transformer had no problem in June or now. He subjects his line to a load test and finds no problem. But, he admits, since I'm still there waiting to see what he'll do, he'll keep looking. He redoes connections atop the meter pole. I'm still there. So he goes up near the transformer. I have to get on to the next job, so I leave. That night my customer calls. He says they found a neutral wiggling in its crimp connector. Wrong size maybe. They gave him the old connector. He'll see if his insurance or the power company will still refuse to reimburse him with that to wave in their faces.

  What easy job was I leaving that one for? Not. Outlets and lights dead in parts of a living room, bathroom and a bedroom. In the living room they read "open neutral". In the bath/bedroom just "dead". Hmmm. It did turn out to be two problems on the same circuit. I found the bad neutral in the living room. A melted wirenut (from aluminum wire). Fine, but now the bath/bed area is still dead. Is it on the same circuit? Only my wire tracer knew for sure. Running signal back on the now-good neutral showed it was the same circuit path as the living room. Where, oh where is the hot bad? Well, after checking at a couple of the first dead things, I went in the attic. The line there was dead from the living room outlet where I had just fixed the neutral! Opened that back up. The black wires weren't melty but they weren't connected well. Now they are. Case closed. In the trenches.

November 10, 2006

  First job was a bedroom and hall out. Plug tester said nothing was hot. Wiggled some of the dead outlets to see if I could get a little life back. No show. The lady pointed me to the panel. Everything looked on there. But my fingers got one breaker to move a little out of ON. Too easily. So I pushed it way off, then hard on. Now she was happy. Her daughter had been using a space heater.

  Next job up a mountain. Vacant house had an airplane's view. Kitchen lights were dead. No life at the switch. The landlady had her handyman show me where the panel was. Two breakers sat tripped. Must have been a little more was dead than she knew. Then she wanted me to hook up a cooktop that had replaced the old one. I expected a small can of worms. I got a big one. The hookup actually looked easy. But why was the line dead when all the breakers were on? She said the old one may not have worked for awhile either. Back at the panel I looked for a cooktop-sized cable and found one. It wasn't in the panel. Had been pulled out and replaced by some other cable. I later traced that to an illegal generator interface, pulled it out, and put our cooktop one in. Some tenant who eats out must have disliked mountain power irregularities. But the story goes on. Putting our cable back in, I found water sitting on and around the main breaker. Not healthy. Sopped that up fine, but the ugly thing I found was all the neutrals on one side were toasted. It was the side used by whoever wired the newer parts of the house. They had a light touch with the screwdriver. She said the house would be torn down in a year or two anyway. I said the circuits wouldn't last that long with a tenant using things. So she had me do what I had to do to restore solid connections.

  Last job today was a few things dead after the guy says he was cutting a lampcord shorter. With it still plugged in. Yes, live. He admitted to being not all there. Now this lamp was plugged in near a GFI in the kitchen. The GFI wouldn't reset. Don't let that fool you. For one thing, it was a newer GFI. It wasn't tripping, just wouldn't catch hold when you try to reset. That means no power to it. For another thing, the cord he cut was just two-wire. That doesn't bother a GFI. But it should have bothered the breaker. Show me the panel! Again things looked OK on first glance. But that double-pole 20-amp was a little off. Why a double pole for two circuits? There can be reasons. Anyway, I reset that. Looks like it was a day for tripped breakers. In the trenches.

November 13, 2006

  (This one may not be of interest to any but electricians). The memorable job today was at an early 1970s house. It had suffered a couple of remodels. A bedroom and bathroom were dead (no hot). After peeking and poking around a little, I resorted to my wire tracer to see where the (good) neutral went. That might point me to the right thing to tear apart. But I never got to trace that neutral. Here's why. The last electrical remodel was recent enough that they had sold the owner a safety package deal. Not only did they put in a new panel, they loaded it with half a dozen arc-fault breakers. To get a signal onto the neutral I have to bring a hot from elsewhere. Usually it is handy to grab that from a good outlet away from the outage area. When I tried that, an arc-fault breaker tripped because they have GFCI sensitivity built into them (for some reason). My tracer wanted to use current along a hot without sending it back on the same neutral. I tried a good outlet on another circuit. It tripped another AFCI. The whole place seemed to be either AFCI or GFCI protected! That forced me to open the panel and get power from a rare normal breaker. Well, an AFCI still tripped! Why? Because the circuit whose neutral I wanted to trace was AFCI protected too, and once a little current started flowing through my tracer onto that neutral (without coming from that circuit's hot), the GFCI feature foiled me again. But all was not lost. At least this finally told me which circuit my outage was on (you can't trust labels). So then I could turn the circuit off and see that some working living room and hall lights and outlets were part of the circuit. I got acquainted with how few connections these working things had and how few the dead ones had. Looks like I have to bite the bullet and go up in that 3/12 pitch attic. With the circuit on, I crawled the length of the attic to where my non-contact tester could follow the live cables to where they went dead. It wasn't clear among the half-dozen junction boxes I found up there (from a 1980s redo) which one was the most likely culprit. When I hit those J-boxes, none of them set off the loud buzzer I had plugged in down in the dead bedroom below. So I guessed and opened a small round box that had 5 12-2 cables entering it. Poking at the black splice there set off my buzzer. I don't usually do this, but I killed the circuit right there by faulting the hot to ground. Who wants to crawl up here more than once? With some difficulty I got the splice redone. The lesson from this house? The new liability of not being able to trace a dead AFCI circuit (without replacing its breaker temporarily with a normal breaker) is offset by the new ability to easily identify the dead circuit because it will trip when its neutral is given current. In the trenches.

September 5, 2008

  This trouble ended up being discovered by my sense of touch. A short circuit tripping the 50-amp breaker for the oven mounted into kitchen cabinets. Appliance guy had already disconnected the oven and the short remained. The homeowner was helpful all along the way. There had been at least one remodel. An older location of the oven was below the gas cooktop across the room. The splices there were not going to be easy to undo and check. So first I used a tracer to follow the oven cable (aluminum) from the panel in the garage into the kitchen. It passed by both of these oven locations and went down a kitchen wall further on, as if there had been a third location. Pulling some cabinet drawers out of the way, I found this third splice box. My testers were not conclusive about which of the three lengths of cable had the short. I gave in and undid all the splices. The short was on the most recently installed cable feeding the current oven. It was a copper cable -- 6-3 Romex with ground. The short proved to be between the red wire and the bare ground. But it only shorted when that ground was in the neutral splice at the preceding junction box. So I was planning to solve the short by taping up this ground at both ends and not connecting it to anything. (Since the cable from the panel had only two hots and a neutral, this ground was redundant anyway.) But the homeowner and I still wondered where this contact between the red and the bare ground was happening. I had already been down in the crawl space under the kitchen once and seen how the two final pieces of oven cable had been run there. I hadn't seen any cable staples being too tight or accidentally penetrating the cable. But I went down under again. Finally I was feeling along the backside of the cable and noticed a rough spot. Turning it toward me I could only see a gleam of metal at one pinpoint. There was no nail coming at the cable from the joist it had been held against, so what was going on? I slit the sheath open and found, amid soot, the ground wire welded (by the short) to a red wire that was skinned bare for several inches. Around the four wires was also a single wrap of electrical tape. The only way you can have tape and a skinned wire inside the sheath of a cable is a mishap at the factory that made the cable. Something during its assembly had scraped the red wire, a worker had noticed and hoped a bit of tape would either alert someone else or keep the wires from shifting before any quality-control testing would reject the reel of cable. The oven at this house had worked for a year since the remodel had installed that cable -- until the red wire finally decided to arc and short to the ground! Factory sloppiness is quite rare. I put a fourth junction box in there in the crawl space as my repair. Literally, in the trenches.

March 28, 2011

  One unit of a condominium complex had a short on a general purpose circuit. Not long ago the management had had an electrician make new connections for all the aluminum wiring in the complex. This circuit fell in that category. Had the electrician done something poorly? And why was a short showing up now, months later. I checked that the breaker was still shorting as reported, even when all cords were unplugged and lights turned off on this circuit. A current tracer I have pointed to one outlet late in the circuit. Removing the receptacle, I saw nothing unusual. So I unconnected the wires and confirmed with an ohmmeter that the short was present between hot and ground on one of the two cables that had attached to this receptacle. On the chance that the shorting cable fed on to the last outlet on the circuit I took the cover off that one. There in plain sight was a hot side-screw terminal of the receptacle right up against the metal box. Yes, shorted. The screw had never been tightened down; it hadn't needed to be, since this outlet was at the end of the circuit and only had one black wire to attach to that side of the receptacle; the second screw on that side was just as it had left the factory -- not screwed down. Even though many people put tape around receptacle and switch screws to prevent this kind of thing, I had never actually seen this happen before. The physical positioning of a receptacle in a metal box wouldn't normally let even a protruding screw touch the side of the box. Why this time? Here is my answer. I saw that the screw seemed more unscrewed than usual. Then I realized that this receptacle was not normal. It was one made to be able to have aluminum wires attach directly to it. Because many aluminum wires are 10 gauge (for 20-amp circuit), these receptacles are designed so their screws back off enough to fit 10 gauge wires under them. Standard receptacles are made with 12 gauge maximum in mind. And why had this screw waited months to short against the box? You tell me...