# When I turn down the dimmer switch on my lights, do I actually use less electricity?

My wife and I have several lights on dimmer switches throughout our home. We generally prefer the ambient lighting provided by dimmed lights. I am curious, though, if we are saving any electricity by dimming the lights. I feel like I read somewhere that dimmers work by rapidly turning on and off the current to a light, though it's just as likely I fundamentally don't understand how a dimmer switch works.

• If you prefer dimmed lights and would never want the lights to be brighter you could always install lower output bulbs instead.
– ChrisF
Jan 26, 2012 at 10:17
• It is better to have less light on, then all the lights dimmed. Jan 26, 2012 at 12:57
• For both answers: It's not true that a simple resistor does not save power. As Power = Voltage² / Resistance, and Voltage is always 230V (or 110V depending on country), the consumed power actually drops. Jan 26, 2012 at 13:45
• I guess @Nikodemus comment comes from eliminating `I` (current) in `P = V*I` and `V = I*R`. But to understand this better it helps me to think about how, as the resistance is increased, current (and hence power) must drop, because the job of mains is to keep `V` from sagging at all under load.
– wim
Jan 26, 2012 at 14:02
• I love this question, and I have a brand new answer that uses thermal infrared data instead of theoretical arguments. I need 9 more reputation before I can post it, but check back soon !! Sep 3, 2022 at 7:03

Yes. And here's why.

# Rheostat dimmers

Old dimmers, used a variable resister to dim the light. Lets look at a simple example.

We can find total resistance (RT), by adding up all the resistance.

RT = R1 + R2 = 0 Ohms + 144 Ohms = 144 Ohms

Then we can find the total current (IT).

IT = ET / RT = 120V / 144 Ohms = .83A

We'll then calculate the voltage across each resistive load.

E1 = IT * R1 = .83A * 0 Ohms = 0V

E2 = IT * R2 = .83A * 144 Ohms = 120V

Finally, we'll calculate the total wattage (WT)

WT = V^2/R = 120V ^2 / 144 Ohms = 100 Watts

Lets see what happens when we increase the resistance of R1

RT = 200 Ohms + 144 Ohms = 344 Ohms

IT = 120V / 344 Ohms = .349A

E1 = .349A * 200 Ohms = 69.77V

E2 = .349A * 144 Ohms = 50.23V

WT = 120V ^2 / 344 = 41.86 Watts

As you can see, we've increased the resistance of R1 and effectively reduced the voltage across R2. And now we have a dim light.

# Thyristor dimmer

Modern dimmers use a TRIAC, to reduce the amount of time the light is on. However, because of the circuitry in the dimmer, there is not a direct 1:1 energy savings. Dimming the light to 50%, will not equate to a 50% savings in electricity.

A typical waveform in an AC system would look like this.

A TRIAC prevents electricity from flowing every time voltage reaches 0, something like this.

So you end up with a waveform that looks like this.

With the TRIAC, the light is actually turning off and on 120 times per second. With every cycle, you're saving a small amount of power. Is it enough to actually see on your electric bill? I guess it would depend on how long the lights are on, and what percentage they are dimmed.

• Some of the newer ones have nice fancy PWM circuits that can start/stop the flow thousands of times a second. Jan 27, 2012 at 13:16
• Yes, those are generally for CFL/LED light bulbs, which take advantage of the circuit design of the bulbs themselves to allow the bulb to dim when a TRIAC wouldn't "trip" the bulb to turn off, and wouldn't provide the "spike" needed to charge the ballast of a CFL to turn it back on. An incandescent normally couldn't care less how you turned it on or off; it responds more to the RMS power in the line than the exact on-off pattern. Jan 27, 2012 at 19:07
• Even for a pure rheostat, increasing the total resistance lowers the total power delivered. P = V^2 / R.
Nov 6, 2012 at 21:51
• Many dimmers use scr's cheaper than triacs and they advertise a 50% savings because they only use the positive cycle (or negative) and dim similar to a triac from there. Basicly a triac is 2 scr's facing opposite directions in parallel. Nov 15, 2017 at 0:32
• Excellent answer, I like the equations. I have a brand new answer that uses thermal infrared data instead of theoretical arguments. I need 9 more reputation before I can post it, but check back soon !! Sep 3, 2022 at 7:04

Short answer is yes, you will save in electrical cost. Probably any dimmer made in the past 20 year has the technology to save you money. This is from Lutron, one of the largest dimmer manufactures in the world.

As you see, not only will you save electricity but your lamps will last longer. Thats why 130 volt lamps last longer than 120 volt lamps.

Dimming LED are easy but to get the best dimming experience you will need a dimmer designed for LED's. These have what I used to call a trim screw so you can adjust the dimmer to use the whole dimming range. Trim screws were used for fan speed controls and you would adjust the trim screw down to where the fan is spinning when the speed control is turned to the lower setting.

• Nice table, this directly addresses the question of whether electricity is saved or not. Sep 3, 2022 at 7:06

While rheostats were used as dimmers in theatrical lighting back in the early days, this has not been common since the 1950s or earlier and I have never seen a domestic light dimmer that didn't use a thyristor. You don't save a lot of energy dimming incandescent lamps due to the highly nonlinear relationship of efficiency to filament temperature (brightness) but it is NOT due to the extra voltage being burned up in the dimmer. Rather as the lamp is dimmed, the light output shifts towards infrared with a larger percentage of the power turning straight into heat in the bulb than visible light. You still save some, but not as much as you might think.

Some of the modern LED bulbs that can be dimmed really do save a lot of power though. On several I've measured, a 10-13W "bulb" dimmed down to what looks roughly half as bright to my eyeballs draws only 2-3 Watts.

• Most theaters used variable transformers or variacs they look like a big wire wound resistor but are not. Nov 15, 2017 at 0:35

I spent a fair amount of time researching this exact question recently, including paying an electrician to come to our home. He had NO understanding of the problem. Most dimmers that you buy are simply variable resistors. This means that if you have a 100 watt bulb on the circuit, but dimmed half way, you are sending out 50 watts to the bulb and 50 watts gets turned into heat in the switch box.

Dump too much heat into the switch box, and you may find you are cooking the dimmer. In our case, 300 watts of bulbs on a dimmer, dimmed down to a low ambient lighting was sufficient to cook a dimmer switch that was rated to handle 500-600 watts. (Our electrician saw that the dimmer switch was theoretically rated to handle the wattage, so it could not possibly be our problem.)

So, no, you are NOT saving electricity at all by dimming a bulb down, at least with a standard dimmer. You can buy LED bulbs, or CFL bulbs to help here. But beware that all LED bulbs do not seem to work on all dimmer switches. And CFL bulbs do not dim terribly well at all, even those that are designed to dim.

You can also buy an electronic dimmer. This is a dimmer that does its job by cutting the power off completely, many times per second. It does indeed save electricity, because the electrons which do not pass on to the light are not just shunted through a resistor to generate heat. Electronic dimmers are more expensive. Note that most dimmer switches you buy at the home store are still the resistor kind.

Finally, you can do one other thing. If you normally run the switch dimmed down quite far, then put fewer or smaller bulbs in the receptacles. For example, we had five 60 watt incandescent bulbs on a single circuit, that we normally ran dimmed way down for ambient lighting. While I plan on buying LED bulbs to replace them, dimmable LEDs are far too expensive now to justify this. Simpler was just to back out 3 of the 5 bulbs. Two 60 watt bulbs, still dimmed down half way are entirely adequate to light the area as we wanted it to be lit.

• To work the math - on a 120 V system, a 100 W bulb is 144 Ω. For the bulb to dissipate 50 W, the voltage across it must be 85 V. This means the voltage across the resistor is 35 V, which means the resistor itself must be 60 Ω, which means the resistor dissipates 20 W. This may be affected by the fact that the light bulb will run at a lower temperature. (and a 100 W bulb may have to dissipate a different amount than 50 W to match the visible light brightness of a 50 W bulb) Jan 26, 2012 at 14:59
• The basic point, though, is - as Nikodemus mentioned - that the overall lightbulb + dimmer system has a higher resistance than the lightbulb alone, and thus has less total current/power. (For an extreme example, for the light bulb to get 0 W of power, the resistor must have an infinite resistance, and therefore also dissipates no power) Jan 26, 2012 at 15:02
• If the bulbs have a weak buzzing sound, then you have a non-resistor dimmer, operating at the line frequency, chopping at 100 or 120 Hz. Jan 26, 2012 at 20:30
• @woodchips: Are you saying that if I put a 1000W lamp a dimmer, and turn it all the way down to the lowest setting, the dimmer is putting out 1000W of heat? Feb 1, 2012 at 17:35
• Every dimmer I have bought is the "chopper" type in one form or another. I can hear the "buzz" or "singing" in the bulb sometimes. Mar 15, 2013 at 6:26

With incandescents (which are, as others have noticed, the only reliably dimmable bulbs), even in the best case scenario, their light output goes down faster than their power usage - so for example (made up numbers, but the principle holds), if you dim them to 75% of normal brightness, you're still using maybe, 80-90% of the original power. The brighter they are, the more efficient they are.

• How does a bulb designed for lower wattage compare to a higher-wattage bulb dimmed to the same light level? Jan 26, 2012 at 19:03
• A 100W bulb dimmed to be as bright as a 60W will use more energy than a 60W bulb running at full brightness. Jan 26, 2012 at 19:59
• Bulbs that operate with that more yellow color are less efficient. That color is the tell-tale sign they are emitting a greater portion of their output in the useless infrared range (unless you are using them for the heat). Jan 26, 2012 at 20:32
• @Skaperen, or IR photography :) Jan 26, 2012 at 21:33
• I know it's like 10 years after you wrote this, but my LED recessed lights are dimming pretty well (flickers only a few times a month). Once my reputation gets high enough I will show an infrared picture of them !! Sep 3, 2022 at 7:11

While it is true that adding a variable resistance in series with a light bulb will lower the current and therefore lower the power (as the resistance increases) the fact remains that power is wasted in the form of heat through the variable resistor. I don't think there are any of these older "rheostat" (variable resistor) types being sold today. The newer design on the market modifies the AC waveform to be on only a portion of each cycle. This design is more efficient as it does not waste unused power however the silicon component used in the dimmer must also dissipate heat and does this through its mounting flange (usually aluminum). This is one reason why only a certain number of switches and wires are permitted to be in the box.

• They used variacs not resistors they look like resistors but are not. Nov 15, 2017 at 0:36

I tested a \$5 Leviton rotary dimmer, rated 600 watts, by dimming Christmas lights. The load totaled 520 watts. The dimmer worked by attenuating the AC voltage supplied. What I found is the dimmer temperature increased when the AC voltage is maxed and the temperature decreased when the voltage was attenuated. I initially thought the heat dissipated goes up when dimming, but now it looks like the majority of the heat generated is due to the inefficiency of the transistors inside. The higher the voltage and more current flowing through the hotter the unit got. At 520 watts load it got hot enough that I could not touch the heatsink longer than a few seconds.

The lesson here is don't use dimmers if you're gonna leave the light on at max most of the time. Energy not converted to light is wasted as heat. Dimming the light will use less electricity and save you money. Whether the light is dimmed or not the dimmer wastes some energy in the form of heat. The wasted energy goes up proportionally with the load.

Dimming your lights will reduce power use. Some people have been suggesting that any power not going out the bulb is instead making heat at the switch. While some power will make heat, you will not be using nearly as much power as at the bulb.

Here is some simple electric math Power Equations: P=IE Power is equal to current times voltage total voltage across the circuit is essentially constant ~120V AC.

If your dimmer is a simple Rheostat resistor, as you raise resistance your current will lower according to Ohms Law V=IR Voltage is equal to current times resistance, since voltage is constant we can rearrange by dividing both sides by "R" to I=V/R to prove that current lowers as resistance increases with a constant voltage.

another way of writing the power law is: P=(V^2)/R with voltage held constant and resistance increasing the output power will lower. Power has a negative correlation with resistance.

if you had a 100W bulb and dimmed to 50W output, you would NOT be producing 50W of heat at the Dimmer. That would burn your house down.

The other type of dimmer you are likely to see is a TRIAC dimmer. This dimmer essentially turns the power on and off over 100 times per second, thus will use less power as the lights are off a greater amount of time out of every second the more that they are dimmed.

You're really talking about a "Watts vs VA" issue.

Go back and look at Tester101's "triac dimmer" illustration.

• Watts is the power you actually use (excluding the black area under the sine wave).
• VA is the entire sinewave that the generator must generate to create the part you use.

"Power factor" is the difference between the watts you are actually using, and the entire sine wave. And triac dimmers have poor "power factor", which varies based on setting obviously.

So for generating capacity, triac dimming doesn't save that much energy, because either the generator has to create the whole sinewave, or some "power factor correction" needs to be done to redistribute the energy around the full waveform. That is something "rotary converters" were rather good at, entirely as a side-effect: the inertia in the spinning machine served as an "AC capacitor" to time-shift energy to redistribute it around the sinewave. Of course now that would be done electronically.

However, your meter only observes the power you are using, and knows nothing about power factor. So you are getting a bargain, only having to pay for the part of the sinewave you use.

• +1 for an intersting comment, but I am a bit confused. Isn't it true that the part of the sine wave you DON'T use is going to be produced by the electric company whether your lights are on or not? Sep 3, 2022 at 7:14

I am answering this question 10 years later, but with some fun data. Many of the answers focus on energy wasted as heat, especially at the dimmer switch itself. I used a thermal-infrared camera to measure temperatures at my dimmer switch. In each panel, the lower temperature reading is at the switch, and the upper temperature reading is on the wall next to it for reference.

The dimmer switch "wastes" roughly the same energy (as heat) on high or low settings, but the light "wastes" much more energy as heat than the switch. So dimming the light saves energy.

Here are some details from the data:

• Even with the switch off, some heat is generated. This seems to be because of the locator light that is always on by default.
• The dimmer switch seemed to heat up one degree more on the high setting. I'm not sure how accurate the thermal camera is.
• The light itself produces a lot more heat, and the switch controls 4 lights. Using the Stefan-Boltzmann law, I estimate that with the dimmer switch on high (bottom row of images), I'm producing about 18x more power in heat at the LED lights than at the dimmer switch.

There is a lot of (sometimes contradictory) info in the other answers about whether dimmer switches are usually thyristor or rheostat type. My particular switch is a Leviton, and I can't find the type on their website, but I get the general impression that those saying that rheostats are not used any more are correct. In any case, the Leviton website gave the following helpful table that also answers the original question for all types of bulbs:

• As much as infrared is super cool, for this case there is a solid demonstration, on visible light, why dimmed incandescent bulbs won't save a thing, or actually even "waste" energy: Turn your dimmer to near minimum. We know that dimmers can't go below 50% by design constraints. But still you can hardly see any light. Where does the remaining 50% energy go to? Correct! Heat (ahm, invisible infrared). Feb 5 at 11:29

Electricity to me is a black science and I have no expertise on the subject. But it seems to me that when using a resistor type dimmer you are pulling power through the meter to the switch and then restricting delivery to the appliance or globe by changing some of that unwanted power to heat. That would mean saving power by using a resistor type dimmer switch is an illusion. Sure the result you see is less light. But what you pay for is not what you see but what is registered at the meter. Newer dimmers that turn the power on/off 120 times a second are a different story.

• Resistor dimmers are antiques -- any dimmer you're going to run into these days is going to be a triac type (or use some equivalent control means) Nov 3, 2017 at 11:41