We bought a house built in 1990 in CA and have been having issues w/ lights flickering when other appliances/fixtures are run. When I turn the powder room fan OFF, a lamp connected to a receptacle 20 feet away in my kitchen flickers. Today, I had a fan connected to the same receptacle with the lamp, and when I shut that fan off it flickered again. Upstairs, when an A/C cycles on in one of the bedrooms, the lights in a completely separate bathroom flicker.

On top of this, single outlets in 2-outlet receptacles randomly stop and start working. One day I'll go to plug something in and one of the outlets don't work. We had an electrician over from the home insurance the sellers left, who used some device on the flickering receptacles and said nothing was wrong and it was "normal."

When I was flipping circuit breakers to match them to my kitchen outlets, I found that ALL the kitchen receptacles (6) were attached to one breaker -- titled "Kitchen." But when I tried to trip that by running 6000+ watts of appliances across receptacles, it didn't trip.

  • 3
    How'd you get 6000 watts of appliances? Was it a single or double breaker? Can you provide a picture of your panel?
    – JACK
    Aug 2, 2020 at 14:48
  • 2
    You might want to pull out one of the bad receptacles and see if the wires are fastened with the screw terminals or pushed into holes in the back. If they are pushed in you may want to redo them and fasten the wires using the screw terminals. In fact you may want to redo all your outlets if they used the backstabs. Aug 2, 2020 at 15:27
  • Is that lamp LED? I've seen LEDs flashing if a TL or other reactive appliance is shut off simply by the apparent power streaking through the house. It wasn't exactly as it should be, but it wasn't problematic either. The houses where that happened haven't burned down and the LEDs can handle it. Do you have a picture of your breaker box?
    – Mast
    Aug 3, 2020 at 6:20

3 Answers 3


It's not normal, and you probably have loose connections somewhere. As usual in my experience, "Home insurance" looks like a scam that does not solve anything. Send incompetent who sees nothing wrong, laugh all the way to the bank. Won't pay for a competent serviceman, they have their own special people who are selectively blind.

If (as Jack asks in his comment) the Kitchen outlets are on a double breaker (Multiwire Branch Circuit or MWBC - very common USA/Canada for the required two 120V 20A circuits in a kitchen) that will support 4800W - I would question whether your "6000W" worth of appliances was actually drawing 6000W, or merely "nameplate rated" for that much total.

There's often quite a difference between maximum draw and actual draw at a given time.

  • 1
    So true about "home insurance". They know how to plug in a "magic eight ball", That's it +1
    – JACK
    Aug 2, 2020 at 15:26
  • 1
    The only way to be sure of the real current draw is putting an amp clamp style meter over the hots in the panel and get a reading. More info and pictures would be helpful to solve this. There may be more than one thing going on here. Some of the newer LED lights are extremely sensitive to voltage variations. if 14 ga wire on a long run, even a relatively small, sudden change in load can cause a flicker, esp. the cheap LEDs. Aug 2, 2020 at 15:50

This is "you getting discouraged" after the first try

I get it, repairs are frustrating, meeting repairpeople is a disruption to your day, claims are intimidating; they're designed to be, to deter frivolous claims. But you have to do them.

It is the nature of intermittent problems that iteration is required to fix them. Because the problem is intermediate, there is no way to do an immediate test: the first attempted repair must be "given a trial" for several times as long as it typically takes to fail.

The most efficient way to handle this is for the electrician to leave, for you to observe, and then, for you to report.

This is when emotions take over. The customer is annoyed that it wasn't fixed the first time, as if my 2nd paragraph did not apply to them. Declares the last repair person incompetent. Never wants to see them again. Is fed up and frustrated with the process. Etc. etc. blah blah. Direct result: the "report" part never happens, and the iteration is interrupted.

You are here.

So branch it anyway you want: a) ignore it, b) bring the free home warranty person back in (believe me, they hate go-backs too), c) hire your own person at your own expense to start the iteration all over again, or d) do it yourself and start your own iteration.

The risk with "taking it on yourself" is after the first iteration fails, you are now obliged to fire yourself, leave yourself a scathing Yelp review, and never allow yourself back in the house again :)

Most likely this thing is...

Builders love to work FAST. If you've done any skilling up in the electrical department, you know that there are 2 (well, 3) ways to attach a wire to a receptacle:

  • wrap a J-hook around a side screw and tighten all to spec
  • jab a straight wire in a hole and done
  • use $3 devices instead of 50 cent devices, place wire(s) in the 2 holes under a screw, and tighten the screw.

One guess which one builders use. This "backstab" method is notorious for causing exactly this sort of problem - and it's usually an expensive, tedious bug hunt. This is why most of us kill backstab connections on sight and convert them into one of the other two.

The worst of it is, backstabs are un-inspectable. You can't check screw torques or see signs of arcing. For this reason, I like to wrest the wire out of the backstab, pulling firmly while twisting maybe 90 degrees, so I can inspect the wire for any pitting from arcing. Also it preserves wire length, which is precious.

Backstabs do have one virtue: they are hard for the worker to screw up. Screw connections, on the other hand, are very vulnerable to being mis-torqued - usually under-torqued. They set up booths at electrician trade shows, and invited electricians to torque screws to spec (without a torque screwdriver). They also tested electrician's companions (wives, office assistants etc.) Both were awful, but both were equally awful, in same proportion of under/good/over torqued. NEC was revised to require torque screwdrivers.


Gold standard advice: you probably have at least a 50% chance of solving this simply by converting all your connections on receptacles and switches from backstab to one of the other types of connection. However, I must recommend a torque screwdriver, because it would defeat the purpose to add as many bad screw connections as you remove bad backstabs!

It's also possible for "wire nut" connections to be screwed up, but the "wire 5 houses a week" type Romex-spider electricians usually get those right. Just give them a pull test: hold the nut and firmly tug each wire. That all but assures a good connection. If there's tape holding the nut together, that's a crutch for someone who couldn't pass a pull test.

Lastly it's worth checking the breaker connection and the neutral wire connection inside the service panel. That last one blind-sides even the experienced experts (who have already checked everything else).

  • 2
    What torque is required for connection screws on standard receptacles and switches? I have got a wiha adjustable torque screwdriver settable from 7.5 - 20 in-lbs that I bought to install AlumiConn connectors. Aug 2, 2020 at 21:30
  • @JimStewart Great question! I know that some devices specify, but I know that there is a "default" value for different devices that don't specify. It seems that this would make a great Wiki-type Q&A that can be referenced from all over.
    – FreeMan
    Aug 3, 2020 at 13:01

This sounds to me like a problem with a multiwire branch circuit. A multiwire branch circuit is a circuit in which the neutral is shared.

normal multiwire branch circuit

This works because Line A and Line B are 180 degrees out of phase, so the neutral current is within the normal range even though it's doing double duty as a return path for both the upper and lower loads in the circuit.

If there is a missing or faulty neutral connection somewhere, what happens is that things will appear to be normal if there happens to be a balanced load, but the voltages will be either much higher or much lower than normal otherwise. That is, if the upper load is much larger than the lower one, the voltage across the upper load will be lower and the voltage across the lower load will be higher.

multiwire branch circuit with faulty neutral

Multiwire branch circuits are required to be protected by a 2-pole breaker which is a single unit and occupies 2 slots or by a tandem breaker which has two separate single phase breakers which are physically connected together so that if either one trips, both do. If they're on opposite buses, this works because the current carried by the shared neutral only has to carry the difference between the current on the two buses, since they're 180 degrees out of phase and cancel. If they were wired to the same phase, the current carried by the shared neutral would be the sum of the current because they would be exactly in phase and additive. That would dangerously overload the neutral conductor and could start a fire.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.