High-Pitch Sound When Lights Turned On (7/25/2K)


Mike, I had a call from my son today regarding his experience in installing a new light fixture and dimmer switch. The fixture has five, 60-watt incandescent bulbs. When the control is anywhere between OFF and full ON he hears a high-pitched sound which is clearly generated by the light bulbs. The sound level goes down with the number of bulbs removed. Have you heard of this phenomenon?

Dr. Dale Rummer, P.E.

Mike Holt's Comment:

I have the same problem in one of my bathrooms and it has five (3 inch diameter) light bulbs. How many members have this problem? Anybody knows the problem and better yet the solution. Let me know, Mike@mikeholt.com

Response No. 1:

I also had this problem a few years back. After making sure, the dimmer was grounded and was not overloaded. I tried every bulb known to man. Found out that clear bulbs hum more than frosted, but what solved the problem was using halogen bulbs.

From: heidiw@netzero.net

Response No. 2:

I have PAR style down-lights in my TV room controlled by a dimmer, and have also noticed a high-pitched "singing" sound when they are dimmed. I guess I always assumed it was the lamp filaments vibrating, but am not quite sure what the physical-electrical cause of that phenomenon would be.

From: Stauffer, Brooke

Response No. 3:

I have never experienced this phenomenon. How can the filament of an incandescent bulb give off a high-pitched sound? I would try a different type of bulb or a different dimmer as a fixed, or call the manufacture of the devices and ask for their input.

From: Rudy F Rangel

Response No. 4:

The humming that you hear is caused by the chopped waveform that is the result of the dimmer. If this dimmer were a simple rheostat, we would have a simple voltage divider circuit and whatever was not used by the bulbs would be dissipated by the rheostat as heat. No cost savings here because the energy used is the same, it's just being shuttled off to different places. The cost saving of a dimmer is the result of the electronics switching on and off the voltage to the bulbs. The way most dimmers work is by utilizing a 'trigger' voltage. As the sinusoidal input rises, the output of the dimmer is still zero until some 'trigger' value (established by the position of the dimmer handle) is achieved. Once this point is met, the output of the dimmer suddenly matches the input and then follows the input until the point that the 'trigger' is met on the downward slope of the input sinusoid. Once this point is met, the output of the dimmer falls (suddenly) to zero. The output waveform, then, is laden with lots of right angles in the waveform. It is these angles, which cause the problem. Especially in inductive loads like ceiling fans. These right angles represent sudden changes in current. For an inductive load, the voltage across the inductor is proportional to the rate of change of current going through it. This also explains why there is never any humming at full voltage. At full voltage, the output waveform follows the input and is thus a smooth sinusoid.

How then to solve the problem? For ceiling fans, either use the big ugly boxes on the walls (these are auto-transformers that retain a smooth sinusoid output) or simply use the pull chain to vary the speed. For lighting circuits? I have never tried anything for these but the only way of addressing the problem is to 'smooth' out the right angles of the output waveform of the dimmer. The only way to do this is to add some reactive components to the circuit. At this point, the discussion becomes somewhat academic, in my opinion, because of the inadvisability of modifying house wiring with components from Radio Shack. The local fire department may not have the staff to handle the results.

I suppose the way to do this is to start investigating dimmers and their characteristics or possibly, by calling the toll-free numbers and speaking with the customer service department to find out if the dimmers have 'soft' outputs. The best thing would be to ask the company to fax (or e-mail) a graph of the output waveform at different settings.

Now that my curiosity has been 'sparked' (excuse the pun) by all of this, I might do this myself. I will forward any information that I get to you. We are moving into a new house and my wife has mentioned that she might like some dimmers. I have never had dimmers in any of our houses, not only for the reasons mentioned above. Also, because of the color spectrum issue, as the voltage across an incandescent filament drops from the design voltage, the color shifts toward the yellow side of the spectrum. This is why a 40-watt bulb has a different color than a 100-watt bulb, operating at a voltage that will produce 40 watts of output. On the other end of the spectrum (groan) are the photoflood bulbs. The whiter color of these bulbs is a result of over-voltage the filament. Of course, this results in an expected life of about 4 hours for these bulbs. "Electronics is a series of compromises based on performance objectives." - from an Electronics textbook.

From: Stromberg, Eric (ER)

Response No. 5:

Theatrical lighting people call this “lamp sing”! When you dim a lamp with anything but a resistance type dimmer (large coil of magnetic wire with carbon slide that varies voltage, big, ugly, lots of heat, very, very reliable), you are actually chopping up the sine wave of the AC current. Your wall box dimmers, all the way up to large theatrical or architectural systems, turn the power on and off 120 times per second (60 on the + side of the wave, 60 on the - side) which effectively dims the light. It is all based on how long we turn it on, 1/2 of the 1/2 sine =50% output, 1/4 of the 1/2 sine =25% output, etc. This distorts the sine wave pretty bad.

Well, at the light bulb, particularly the more inexpensive household line of lamps, the filament is only supported at the ends. The “Sing” occurs when we turn the power on and off 120 times a second, slamming the filament back and forth from rest to positive to negative to positive.

Some dimmer manufacturers put chokes in their dimmers to help suppress the noise and distortion of the AC wave (inductance at work here, don't ask me to explain that part) and it helps to calm the “sing” down. In addition, better lamps with better filament supports keep it from rocking and rolling.

Solution? Use better lamps and better dimmers. There are some new electronics dimmers out there for all of this, but they are only available in theatre and studio applications now.

From: Bill Ellis

Response No. 6:

He may have used the wrong kind of dimmer, a fan speed controller instead of an incandescent dimmer. Hope this helps.

From: Limbaugh, Margaret

Response No. 7:

I had the problem in the early 80's in a church with 500-watt bulbs. I attributed it to filament vibration.

From: Douglas Michaelson

Response No. 8:

The problem is typical with new dimmers that are not rated for the load that they are trying to dim. I would also check to make sure that the grounds are properly attached to the fixtures and the dimming switch itself and that any metal boxes are properly grounded. Any one of these things has set up nuisance noise problems for me in the past as the lead electrician in a hospital.

From: caj@infocom.com

Response No. 9:

The ear can only detect changes in air pressure. Light dimmers use a silicon-controlled rectifier to turn on power for a part of the sign wave. This produces a pulse 120 times per second, alternating negative, and positive. This pulse must be exciting some component to resonate and vibrate at a high audio frequency.

From: bp6769@aol.com

Response No. 10:

Yes, I have a wall sconce in my house (75 or 60 watt bulb) on a dimmer that does the same thing. I also have one of those popular freestanding brass lamps - the kind on a 5' pole with a halogen bulb, with the dimmer switch as part of the equipment that does the same thing.

From: McKnight, Roger

Response No. 11:

I had something like this happen a number of years ago, my problem turned out to be cheap indoor flood lamps, I replaced the lamps with GE lamps and my high pitch noise from the light cans was resolved. There were six 150 watt indoor floods on the same circuit.

From: Frank, jmelect@aol.com

Response No. 12:

I have encountered this problem before. It is caused by filament vibration when the fixture is dimmed. It has always been in bulbs that have a long vertical filament as opposed to those that have three posts with the filament stretched between them horizontally. Not everyone's hearing can pick it up but for those who do hear it, it can be very annoying. Usually changing the bulbs to either the other configuration or using rough service will take care of it.

From: Kramer, Stephen

Response No. 13:

Several years ago, I had this resonating noise problem when I installed hi-hat fixtures in my kitchen with Lutron dimmers. I called Lutron, and they claim this problem lies with the bulb in that the filament vibrates when the bulbs are dimmed. I do not know the answer to the problem but it appears to sound worse when you have a ceramic tile floor.

From: Bufis, Phil

Response No. 14:

Check out competing brands of dimmers. Some may advertise on the package that they are interference reduced, etc. This noise often referred to, as SCR hash, from the SCR solid-state devices used in the circuitry, often tends to leak into home audio systems, computer gear, and other stuff. Grounding does not change anything, as this is a radio signal (sort of) that travels both acoustically from the vibrating lamp filaments, and electrically through the wires.

Sometimes using heavy-duty industrial strength lamps reduces the acoustic noise level, as the filaments are suspended from heavier supports within the lamp. One thing to verify is a good “third wire ground" within the house system. Aside from the danger of poor grounding, inefficient grounding magnifies the problem. Look for reversed hot and neutral wires in the wiring of the dimmers as well. There is no true cure, other than purchasing better quality dimmers.

From: Soifer, Allan

Response No. 15:

The whine or singing of the lamps is common to either fluorescent or incandescent bulbs. It is caused by the high frequency switching rate of the electronic dimming circuit. This is why the manufacturers of professional dimmers for stage and TV studios go to great lengths to use toroidal transformers, high precision rectifiers, and to send clean DC to the lamp filaments, or filtered AC when alternating current is mandated. That is why the dimmers for homes are a few bucks, and the pro models cost into the thousands $$$.

From: Allan Soifer

Response No. 16:

As it so happens, I have had some experience relating to sound dampening and transmission. It seems to me that the lamp holders are not securely tightened in their metal housing and/or you need to install some kind of sound dampening material, cardboard, insulation, rubber, plastic, (or thermal plastic if so required). In addition, you may want to check for lamp tightness, as this will contribute to resonance transmission too. Also securing the fixture tightly to the surface will diminish the vibration too. All these little things will add up to that irritating hum and/or noise depending on all the factors involved.

Although my experience is related to industrial machinery and environments, these suggestions should apply to this situation as well. If after all else fails, replace your lamps with new domestic (made in USA) ones.

From: Sirjaxx2@aol.com

Response No. 17:

It is due to the construction of the dimmer that the lamps are "singing". In dimmers, the triac or solid-state relay will turn on at different places along the sine wave depending on where the slider is set. At 0% it never turns on (180 degrees phase angle), at 50% it turns on half way through each half cycle (90 degrees phase angle), and at 100% it never turns off, or more accurately it turns on at the very beginning of each half cycle (0 degrees phase angle). This switching gives you a fundamental frequency of 120hz and depending on the load, you will get a wide array of harmonics in different amplitudes.

This set of frequencies makes the filament in your lamps vibrate or "sing". In theatrical dimmers, a choke (ferrite torrid core with a single wire wound around it) from 280 milliseconds to 800 milliseconds removes almost all of the "singing" while not introducing too much inductance in the load. Imagine, 3000 lamps all singing at the same thing in a quiet theatre? In the little wall dimmers, there generally is not enough space to put a choke in the case since a 20A choke would fill a 2G box. You could however put a choke elsewhere inline with the load.

From: Wescoatt, Michael

Response No. 18:

The lamps are humming from the chopped wave produced by the triac for dimming. It is the vibrating of the filament within the lamp, which makes this singing sound.

From: Thompson, Scott

Response No. 19:

Yes, I have had the same singing many times. I believe it is due to half-wave rectification and large long filaments!

From: Walker, Bill

Response No. 20:

I have heard of this happening before and the solution was a different brand of bulb. I don’t know exactly what is happening but I venture to guess that the filament is vibrating because of the way that the triac in the dimmer chops the top of the wave form off to give the dimming effect. The case I heard of was in the remodel of a very expensive home where everything had to be perfect. Changing the bulbs solved the problem.

From: Currin, Richard

Response No. 21:

Reason – Cheaper dimmers chop up the waveform, thus bulb filament is vibrating due to “pulsed” power.

Use high quality dimmer that has internal filter and “tough service” lamps that have filaments that are more rigid. Not practicable, but longer wire adds more inductance to the circuit.

From: Maddox, Robert

Response No. 22:

I had the same problem until I changed the type of dimmer I was using. Try going with a better dimmer.

From: Ressman, Dennis

  Go to top of page
Newsletter Registration   |   Stay Connected:

888.NEC.CODE (632.2633) 3604 PARKWAY BLVD, STE 3, LEESBURG FL 34748  

Tell a Friend About This Site

  NEC® and National Electrical Code® are registered trade marks
of the National Fire Protection Association (NFPA).
  ©Copyright 2011 Mike Holt Enterprises, Inc.