Search This Blog

Wednesday, August 07, 2019

Why do some LED bulbs create radio frequency noise?

A while back, I tried to replace the incandescent bulbs in our garage door openers with LED bulbs. 

I started my next shift and promptly forgot that I'd switched them out - however, the garage doors stopped working with the remote.  My wife was the one who put 2+2 together, and realized that this loss of function was due to my replacing the bulbs.  I switched back to the incandescent bulbs, and the problem vanished.

I then did a little research and learned that this is a pretty common complaint with LED bulbs and garage door remotes.  I've since managed a work-around to get LED lighting, but the problem got me thinking - and back to the title of this post: "Why do LED bulbs create radio frequency noise?".


Well, it's not really the "LED" (Light Emitting Diode) part of the bulb that is causing the problem.  LEDs are simple devices that take a small DC current and convert it to light, with pretty high efficiency.  Consistent DC current flow doesn't cause radio frequency (RF) interference.  The LED in your flashlight uses batteries, and won't make any RF racket.

To get an understanding about radio-frequency interference that some of these bulbs emit, we need to discuss how we power up LEDs from household 120 volt AC.  Let's see how we go about changing that high voltage AC power to low voltage DC power to get a better sense of what happens in a LED bulb.

There is another type of diode - one that does not emit light.  This diode is used only to allow electric current to flow in one direction.

In a water system, the mechanical equivalent of a diode would be a check valve, below.  In the image below, water can flow from left to right, but not from right to left.  That is because the flapper will shut, blocking reverse flow.  Diodes behave in a similar manner with respect to electron flow, by using a semiconductor junction.







Below, a high-magnification image of a diode.  Image courtesy of  John Maushammer.

In a house, all of the light bulb sockets, and most of the electrical power outlets are energized with 120 volt, 60 Hz AC power.  To light a single LED requires a forward bias voltage of 3-5 volts DC, depending on the type and color output of the LED.

If we are going to use a LED, we first need to convert the AC power to DC power, and then we need to reduce the voltage so that the LED won't explode.  Alternatively, we can string 24 to 40 LEDs in series so that the 3-5 volts add up to 120 volts.

There is a third thing we need to do electrically that is not obvious at first glance.  We need to limit the current that can flow through our string of light-emitting diodes.

In an ideal world, resistance to forward current flow in a diode would be zero, and resistance to reverse current flow would be infinite.  We don't live in that world.  However, forward biased diodes don't have very much resistance to flow, and so they will burn up due to internal heating if you don't limit current somehow.  Usually a resistor is placed in the circuit to prevent runaway current flow.

Now that we have discussed diodes and their little quirks, let's see if we can make a basic power supply to drive a LED bulb and see why it makes RF noise.

Let's first convert 120 Volt AC power to DC power.  To accomplish this, we have to connect a few regular (not light-emitting) diodes together in a certain way.  The diodes are connected to allow current to flow when the applied AC is either positive or negative, but "flips" the negative portion of the sine wave to positive.

In the image below, an AC source voltage is applied on the left.  The voltmeter shows a sine wave with both positive and negative voltages.   This voltage passes through the four diodes of the full wave bridge rectifier.  Depending on the biasing of each diode, it will either pass or block current flow.  Each leg of the bridge either passes or blocks current flow depending on whether the diode is biased to conduct or to block flow.  The output of the bridge on the right is a pulsed DC voltage, with twice the peak-to-peak frequency of the source voltage.




So here is one minor source of noise.  Electrical switching from the diodes and 120 Hz DC ripple.  You also get harmonics (multiples) of the 120 Hz ripple.  The ripple is removed by a filter capacitor.  This capacitor helps to maintain a consistent voltage as the DC ripple across the load drops to zero.  There is a great video explaining this process below:

Sometimes a solid state voltage regulator is also used to maintain the DC voltage.  This shunts off voltage above a certain value, and also creates electronic noise due to switching on and off several times per second.

Some LED bulbs contain a switched mode power supply (SMPS).  A SMPS can be made small enough to fit inside the base of a light bulb.  Its purpose is to change the DC voltage from the bridge rectifier to a higher or lower voltage.  The SMPS This is probably the most significant source of RF noise that a bulb gives off, and that's due to the fact that the power supply switches polarity thousands of times per second.

Below, a basic block diagram of a Switched Mode Power Supply.  The left block is the bridge rectifier and filter capacitor discussed above.  The next block to the right is a high frequency switching circuit that turns on and off thousands of times per second.  Next over is a tiny transformer that provides the desired output voltage, followed by another bridge rectifier to convert the new voltage back to DC.  The lower blocks are feedback to maintain the desired output voltage as the circuit is loaded and unloaded.  Yes, it tends to make a lot of RF racket.

Below is a fellow who is checking a switch mode power supply for radio frequency noise.  He is more concerned about how to get a correct reading, but it also shows how much RF noise these high frequency switching power supplies make.


If you are curious about how much radio frequency noise a generic LED bulb emits, here is a cool link to an article showing a spectrum analyzer connected to a radio antenna with a few different bulbs. 

No comments: