Hcfyk

The wiring in my house is... odd. Nearly every pair of outlets has one of the outlets connected to a light switch. Sometimes that switch is quite far from the outlet even when there are other closer switches that seem to control nothing at all. Sometimes that switch also controls overhead lighting, which makes no sense to me from a design standpoint. Sometimes that switch is a dimmer switch, which is a really horrible idea as if you plug anything other than a light bulb into it, like some electronics, the dimmer switch could do them a lot of damage. One such dimmer switch is also connected to a ceiling fan, and setting the switch to anything less than full does not change the speed of the fan, it just makes it a lot noisier (AC induction motor, obviously).

There are light fixtures that don't seem to have any power to them and aren't controlled by any switch, though I could believe the switch might just be hiding somewhere. There are also switches that seem to control absolutely nothing, though I could believe I just haven't found what they're connected to.

The oddest thing I have found though has to do with the circuit breakers around overhead lighting. I have a chandelier in my dining room that is connected to a dimmer switch. When I moved in, all of the bulbs were incandescents, and I wanted to be more energy efficient, so I replaced them all with dimmable LED bulbs. Now however, I could not turn the chandelier fully off, with the dimmer switch all the way down, the lights still dimly shone. This had never been an issue before, but I thought it probably meant the dimmer switch was faulty, and it was just the different voltage response curves of the incandescents versus the LEDs that meant it had never been noticed before, so I set out to replace the switch.

Now there are two circuit breakers that are relevant to this chandelier. One is labeled "dining room," the other is labeled "lights." As far as I can tell, the dining room breaker controls all of the outlets in the dining room area, while the lights breaker controls all of the overhead lighting in the house.

Here's where the odd part comes in. If I turn off the dining room breaker, the lights stay lit. I have turn off the lights breaker to switch off the lights. However, if I only turn off the lights breaker, and go into the switch panel with my voltmeter, I am still detecting 100V AC across the switch (with it off), even though the lights stay off when I turn the switch on. In order to remove the power I was detecting, I have to also turn off the dining room breaker, even though the lights were staying on when I switched that off.

This conforms to no type of electrical theory with which I am familiar, though there might be some peculiarity of household wiring I don't know about that explains this. Can anyone tell me what might be going on?

Update 3/24/19: After replacing the dining room dimmer switch with one specifically designed for LEDs, those lights are now working quite well. Based on several people's advice in this question, I have now set out to replace the bedroom dimmer switches that are connected to outlets with toggle switches. Have run into additional issues detailed here.

However, if I only turn off the lights breaker, and go into the switch panel with my voltmeter, I am still detecting 100V AC across the switch (with it off), even though the lights stay off when I turn the switch on. … Can anyone tell me what might be going on?

Coupling between wires (inductance and capacitance in parallel wires).

You say you have strange long runs between outlets and switches — if any of these extra lengths run in parallel, this is a great environment for the voltage from one to be induced in the other if it's disconnected.

If you try putting a load, like a small light bulb, across that 100 V, it will vanish (read 0 V), because there is no significant current capacity in this coupling. It might even go away if you were to attach an analog voltmeter (which is more of a load because current from the line to move the needle) instead of a digital one (which typically has a 1 MΩ input impedance that presents almost no load).

Rundown in layman's terms:

1. Light switch to outlet. You will see a lot of these in older homes. By code for a long time in a lot of areas you had to be able to turn on a light from a switch accessible when entering a room. Easiest and most cost-effective way to do this is put switch to outlet. I had six rooms in my house like this.




2. Dimmer to outlet. For sure not code compliant. You are correct that it could damage electronics and could be a huge fire/safety issue. Please replace this with a regular switch ASAP - $5 for a quick fix while you work things out.




3. Dimmer to ceiling fan. Can't do this. If the ceiling fan has two hots coming into the box from fan you can buy a fan control + dimmer. If there is only one hot then you just install a switch and it is off/on for both.




4. Light switches doing nothing. You must do a complete mapping assessment of your house. With a house like this it might take 3-6 hours. But you have to turn off one circuit at a time and figure out what doesn't work. You must also for 100% sure make sure that each switch is tested for working on every light and outlet (have a known working bulb). Write down every outlet and switch in a spreadsheet marking room, location, what it controls, and circuit.




5. You cannot expect to plug LED bulbs into old dimmers and expect them to work. You need to put in a dimmer that is suitable for the type of LED you are installing.




6. The dining room lights - need more info.

Switch switching outlets. That's normal enough. Often the outlet(s) switched are far from the switch, because that is where the installer thought you would want to put the light. The installer was wrong. They always are. Because of that some of them switch more than one receptacle.

Split receptacles. That is normal. That's what the "tab" between the brass screws is for. Leave the tab on, the 2 screws are an easy way to splice. Break it off, you control each socket with a separate hot. There's also a tab on the neutral side if you want the sockets on 2 separate circuits.

Dimmers on receptacles. No no no! For the reasons you stated and worse. You know what to do. If you were super attached to dimmable receptacles, they make specially keyed receptacles just for lamps.

Mystery lamps. Fair possibility the bulbs are burnt out (back in the old days, bulbs burned out, true fact) or there is something wrong with the fixture. They make wire tracing devices if you really want to dive into the mysteries.

Mystery switches. Aside from that they may control a broken lamp, sometimes that just happens as wiring gets revised. They may also control receptacles that you have not checked, outside lights you do not see (optionally: with burned out bulbs) etc. They may also be half of a miswired 3-way circuit that is miswired in a way only the other switch works. **3-way switches do not have the words "on" or "off" on them.

Glowing LEDs on dimmers. Actually, the incandescents also glowed, but only in infrared, so you would have needed to be an elf to see it. LEDs are so much more efficient (and glow only in the visible range) that it's much more obvious. The reason is many switch loops do not have neutral wires. That means dimmers cannot power themselves between supply and neutral; they must place themselves in series with the lamps. That works slick on incandescents.

By the way, some dimmers actually have a "hard off" or even "hard on" setting. It's not in the smooth working range of the dimmer. You have to push the knob harder past a detent to reach that position. Most people, it never occurs to them to push it that hard, so they never know it's there. The "hard off" is important so you can change the bulb safely. The "hard on" is nice because it stops the bulb from buzzing. Dimmers aren't variacs, and their waveform is real sawtoothey.

Grouping of lamp circuits. It is common and wise for lamps to be powered from a different circuit than receptacles. That way if you trip a breaker you are not also in the pitch black. It is extremely unusual for lighting to trip a breaker (especially if LED), however they need to be provisioned with a worst case scenario (all oversized incandescents) in mind, and 125% of that.

Crosstalk. This is plain old capacitive coupling. Lay two wires alongside each other, one energized with AC, one disconencted, this will "antenna" some microscopic current onto the other wire. Cheap DVMs are so sensitive they can see this. Use a traditional analog voltmeter and that it will extinguish that phantom voltage; there's not enough current behind the phantom voltage to move the jeweled meter movement.

Multi-wire branch circuits (MWBC). This is a pair of half-circuits wired using a common neutral. The two half-circuits must be 240V across the hots, with neutral in the middle. That way the neutral carries the difference in currents, and you can trust it won't exceed breaker value. Done wrong, neutral carries the sum of currents, which overloads it!

In any circuit, all wires must be carried in the same cable or conduit so currents in each direction cancel each other out. This is AC power: any gap will induce vibration and fatigue failure. All that is relevant because the phantom-voltage wire may well be half of an MWBC, which would explain the strong capacitive coupling.

There are additional rules with MWBCs. Neutrals must be pigtailed, not use a device like a receptacle as a splice point. The breakers must be handle-tied so they shut off together, but normally a 2-pole breaker is used.

I expect your double circuit (the one with the phantom voltage) is a MWBC, and the work was done back before handle ties were required.

MWBCs have fallen out of favor because they are difficult to use with AFCI or GFCI protection, which is now required on almost all circuits.