RF interference from LED bulbs

[GoldDigger]

There have been discussions earlier about RF interference from the driver circuitry in LED bulbs, highly variable between bulb manufacturer and model. The worst case discussed involved the light socket for an A19 bulb found in most garage door openers. The main symptom is that the remote control has a decent range for opening the door, but once the bulb has turned on the control range is reduced, often to uselessness, by the RF output of the bulb interfering with the receiver a few inches away.
The recommendation was always just to try a different bulb type.
Well, today at Big Orange, while looking for a replacement gate remote (they were out of stock!) I noticed that Genie (the garage door opener maker) is offering under their brand an A19 LED bulb which does not cause interference. Reasonably priced at only $10.
Find a need and fill it.....

 

[Little Bill]

I haven't tried LED bulbs in my openers yet. I just stuck in incandescent bulbs because that's all I had at the time. I planned on replacing them with LED once they blow (been in about 2 years). I might just experiment and see if they interfere with the remote before trying the new Genie LED.

 

[synchro]

There have been discussions earlier about RF interference from the driver circuitry in LED bulbs, highly variable between bulb manufacturer and model. The worst case discussed involved the light socket for an A19 bulb found in most garage door openers. The main symptom is that the remote control has a decent range for opening the door, but once the bulb has turned on the control range is reduced, often to uselessness, by the RF output of the bulb interfering with the receiver a few inches away.
The recommendation was always just to try a different bulb type.
Well, today at Big Orange, while looking for a replacement gate remote (they were out of stock!) I noticed that Genie (the garage door opener maker) is offering under their brand an A19 LED bulb which does not cause interference. Reasonably priced at only $10.
Find a need and fill it.....

I wonder if the receiver desensitization from the bulb is mainly due to RF interference conducted through the lamp socket, or if it's primarily from radiated interference that's picked up by the internal antenna. If this problem occurs with a particular bulb then it could be put into a handheld lamp ("trouble light", etc.) and placed in close proximity to the receiver to see if radiated RF is enough to induce the problem. If not then conducted RFI through the socket is more likely to be the dominant mechanism.

Perhaps Genie just tested different LED bulbs that were already being produced and found a suitable one, or they had a manufacturer customize one. Probably the former unless they couldn't find anything. If the problem was mainly from conducted interference they could've put some RF filtering inline with the bulb. But that would have cost some money while selling special bulbs makes more money, so that's a no-brainer!

 

[gaelectric]

This is a good thing to know about. Thanks. I've had trouble explaining this to clients. Nice to have a possible solution.

 

[steve66]

I couldn't get an LED lamp to work at all in my garage door opener. But I only tried one brand. It just wouldn't even turn on, so I assumed the switching circuit in the opener needed the conductivity through a standard lamp to work. Similar to the way some motion sensors won't work with LED.

Too bad because I hate how the heat from even a 60W lamp turns the opener cover yellow.

 

[gar]

191021-2215 EDT

I bought one of the Genie bulbs today.

Ran tests on three different bulbs. Measured the bulb current with a 2.7 ohm carbon composition resistor.

Genie bulb --- no apparent significant high frequency current, but the 60 Hz current is not real good. Note the box does not guarantee that the bulb will not cause RFI.

CREE --- no apparent high frequency current. 60 Hz current is fairly close to a sine wave. From previous tests this bulb dims well with a phase shift dimmer, or Variac. And does not cause RFI.

Feit --- a particular model that does cause major RFI around the middle of the AM broadcast band from previous tests. This oscillation conducts on to the power wires, then radiates from the pole wires to interfere with my car radio as far as 100 ft from the last pole. This about 1 MHz oscillation shows as a major component superimposed on the 2.7 ohm 60 Hz signal in my present test. This oscillation on this particular bulb drops out below about 110 VAC.

The CREE bulb is probably as good or better than the Genie, and a lower cost.

.

 

[GoldDigger]

Thanks Gar, always appreciate some hard experimental results!

 

[gar]

191021-2337 EDT

GoldDigger:

Not all Feit's have a problem. This showed up on a number of samples of one model, a very low cost one at Home Depot. Another model showed no problem.

I can not say that a CREE won't have a problem. I think my scope, and 2.7 ohm carbon comp current shunt are a useful tool for this purpose. So is a portable AM radio.

.

 

[gar]

191022-1308 EDT

Decided to do some other tests.

The test was going to be the radio test. Set up my receiver with AM broadcast, and 3.7 to 30 MHz range. This is being line powered. The test bulbs were not on. Various things in the house were on. On AM I was getting an unwanted noise signal at about 550, 680, 800, 940, 1080, 1200, 1350, and 1600. The 1350 was a different noise than the others. Note the difference frequencies are in the range of 120 to 150 kHz. This looks like a carrier current signal. I don't have a TED system on at the present time. The signal is continuous, but from the speaker it is raspy.

A fundamental in the range of the difference frequencies would appear to be the source.

Turned off virtually anything in the house that might be the source. The signal did not go away.

Next looked at the AC waveform and it looks clean.

Don't know where my tuned filter at 120 kHz is. So I made a simple high pass filter with a 560 pfd, and 1 k resistor. Fed the AC line into this to observe high frequencies on the line. The waveform appears to have two major components with periods of 19 microseconds and its third harmonic 6.33 microseconds. Or frequencies of 53 kHz, and 159 kHz. If I start at 550 and repetitively add 159 I get somewhat of a correlation with the above first stated sequence of frequencies, but not real good. Nowhere as good as I might expect.

Some future time I will go battery powered and look around.

Back to the LED tests later.

.

 

[gar]

191022-2356 EDT

Earlier today I put a capacitor across the phase feeding my bench where I was seeing the high frequency signal described above. No capacitor on the other phase. Both phases feed the bench. The capacitor basically turned the scope trace into a straight line, did not change gain. The other phase still had high frequency on it. Signals were still present on the receiver.

Tonight I did a different sync on the scope. Synced to 60 Hz. Saw high frequency for a portion of each 60 Hz half cycle. Not at zero crossing of 60 Hz voltage. On top of the high frequency oscillation is a 2 kHz modulation. These bursts of high frequency are not precisely synced to the 60 Hz, but float over a small range for their starting point in the 60 Hz cycle. But they are definitely correlated to the 60 Hz signal.

Still no idea of the origin of this interference.

Does anyone else look for unwanted signals on their AC power?

.

 

[ELA]

1

 

[gar]

191023-2406 EDT

ELA:

The 40 kHz is likely the oscillator in the in the bulb.

Can you hear harmonics of this in an AM receiver?

By turning off virtually anything in my house that might originate the raspy noise frequencies I hear on the AM radio does not get rid of them. I need to battery power the receiver and go outside, and see if when I am away from home or my primary wires if there is a change.

.

 

[GoldDigger]

Corona discharge across a faulty insulator on the POCO distribution lines? I guess that would not have a well defined frequency unless there is a resonant circuit involved.

Sent from my XT1585 using Tapatalk

 

[ELA]

Hi Gar,
Sorry but I seldom listen to AM radio. And the bulb was returned long ago.

I have a lot of unwanted frequencies on my power lines.
If I should have to know, I would setup up a Spectrum Analyzer for harmonics, from either radiated or conducted emissions.

 

[gar]

101029-1732 EDT



ELA:

Many LED drivers seem to be in the 40 to 50 kHz range. If these don't go into some spurious oscillation, then I doubt they would interfere with a garage door opener.

More experiments on my 550 kHz noise will follow later.

.

 

[Speedskater]

LED bulbs and lighting systems are a rapidly developing and cost reducing field. A product that was purchased last yeay many have no engineering similarity to a product with the same part & model number purchased this year.