Setup: In a bedroom in my house, The “light” switch by the door controls one of the wall outlets. I recently bought a floor lamp that came with compact florescent bulbs (the helical integrated kind), and plugged that into the switch-controlled outlet. With the standard single-pole wall switch, it worked just fine.
I recently swapped the switch for something similar to this one (I don’t think this is the exact one I bought).
For a while, everything worked fine. But the other day I flipped the switch on and the CFLs in the lamp just flickered, very dim. Not only that, but they continue to flicker (although even dimmer) when I turn the switch OFF.
Huh? Doesn’t hitting the “off” side of the switch completely shut off the flow of electricity to the wall outlet? So how can the bulbs still flicker?
A look at a typical dimmer circuit will show that even when “off” there are current paths through capacitors and resistors (the capacitor marked C1 is for noise suppression due to triac switching, the others shift the phase, along with the resistors; this is for a 240v circuit but 120v models are basically the same). The leakage current may be only several milliamps, but is enough to charge to capacitors inside the CFL, which usually draw no significant current until the voltage gets high enough for the circuitry to operate (for example, one that I opened used a diac to kick-start the inverter, as shown in the typical circuit here, 120v models use a voltage doubler instead of bridge rectifier).
CFLs aren’t inductive loads, unless you have one of those really old (1980s?) models with a magnetic ballast; they are instead complex non-linear loads (some would say capacitive but that isn’t quite true since a rectifier isolates the capacitors and the current waveform is a series of sharp spikes; the problem is that triacs need current to stay on, unless you continuously apply current to the gate, which would require a separate power supply). However, it is true that dimmers shouldn’t be used on inductive loads, unless designed for it.
I started having the same problem and I can’t figure it out. When I was having some work done, the contractor put a new (really cheap) florescent light in the closet to replace one that had a pull chain and installed a switch on the wall. it is a rocker switch that looks like the one the OP referenced, but it is a single pole, plain-old on/off switch. When the florescent burned out, the fixture fell apart as I tried to remove the tube, and I replaced it with a simple screw-in incandescent fixture. I put in a CFL bulb. When the power is off, it flickers. Why doesn’t the single pole on/off switch kill all the power to the light?
Well, thanks for the replies. I’m not sure I understand any better, though.
As I said, the switch I linked to is not the exact one that I bought (I bought it from a hardware store, and I don’t remember the actual brand). I wish I had saved the packaging, but I am pretty sure that it said “CFL compatible” on it. And I know it’s not a dimmer by looking at it, there’s no slider or anything, and multiple pushes of the “on” side do nothing. The only thing special about it (at least from the outside) is that it comes with a separate unit that is an RF switch.
Is it possible that even though the switch itself is not a dimmer, that it has a dimmer circuit inside? As a WAG, perhaps it’s cheaper to manufacture one kind of “innards”, and just package it once in a switch that has the dimmer slider, and again in one that doesn’t? That’s the only thing that seems to explain how the bulbs could flicker when the switch is supposedly off.
traditional dimmers use TRIACs to change the voltage/current seen by the lamp. By doing so, they alter the waveform from the sine wave supplied by the mains power into something that is “not a sine wave.” Incandescent bulbs don’t care, since they’re just a little piece of wire which gets really hot and glows. LED and CFL bulbs have power supplies and ballasts (respectively) and can’t cope with the hacked-up waveform coming from a TRIAC.
To answer your question, yes, the switch you’re using is probably like a dimmer in that it uses TRIAC(s) to do the switching. Photocell switches use TRIACs as well, and if you have an LED bulb and want to put it on a photocell, the photocell needs to be “LED/CFL compatible.”
ETA: I make no claim that this is a good solution. I have no idea.
The only thing I can think of is if the switch is on the neutral side (leaving the bulb “hot” even when off), there could somehow be enough inductance that some current can still flow. If so, if you could rewire to have the switch cut the hot, so the bulb was at neutral when the switch was off, that should fix it.
That would be my guess that they just used a triac for both the dimming and non-dimming versions. The right way to do it, as in my Insteon switches, are to use a triac for dimming and a relay for a non-dimming load.
AFIAK, dimmers aren’t designed “for” LEDs and CFLs, it’s the other way around in some LEDs and CFLs are designed for dimmers, with varying degrees of sucess. Even if a dimmer was marked as such I’d be very hesitant to use it to control a load that wasn’t marked.
Unless you hear a clicking noise in the main switch when you use the associated RF switch, indicating a relay, then the RF switching functionality is probably accomplished by a TRIAC in the main switch.
That wouldn’t cause any current to flow unless there was significant leakage (again, not inductance, don’t know why everybody is saying that); again, the reason is because triacs almost always have a snubber circuit across them (sometimes just a capacitor, or a capacitor and resistor) which causes some current to flow even when it is off (for RFI purposes and to prevent voltage spikes from turning the triac on; the triac itself has almost no leakage when off, certainly not enough to have any effect).
For some calculations, in a previous post I linked to a typical triac-based circuit, which has a 0.1 uF capacitor across the triac, which will let 4.5 mA flow at 120 v (about half a watt, or 2 watts at 240 v, somewhat less at 50 Hz, which sounds wasteful but it doesn’t affect your electric bill since it is reactive power and dimmers, at least one that I have, have a full-off position which actually breaks the circuit).
Also, one reason why a triac would be used and not a relay is that a triac can be turned on with only a brief pulse; it stays on until the current falls to zero (thus, they only work on AC; on DC they will stay on until power is removed; this is also how dimmers can work with just two wires). A relay needs a continuous (and much higher) amount of current to keep it on, which uses a lot more power and needs a dedicated power supply, meaning three wires to the switch (hot, neutral and load). However, relays are more efficient at higher loads since triacs drop 1-2 volts when on (and also don’t need any snubber/noise suppression circuit, thus no leakage).
Another thing- did the switch have a neutral wire? Smart switches have to have some way of operating the electronics inside whether the switch is on or off (if they killed the circuit cold the’d have no way of knowing when a remote signal to turn on was given). Insteon switches simply mandate a neutral wire, but some X10 switches and (presumably) others would supply power to themselves by putting the electronics in series with the load, so the return path for the current used by the electronics is through the “off” light bulb. It’s not enough power to light up an incandescent bulb even barely, but can certainly cause weird things to happen with non-incandescent loads. That’s another reason for not using a relay- no point in it if it can’t be used with non-incandescent lamps anyway and as has been speculated they could probably use one PC board for both dimming and non-dimming versions that way.