Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Bill Meeks]

Thought I would share this with the group in case anyone else runs into a similar issue.  If you Google "LED RF interference" you will find other stories of RF interference that eminates from the switching power supplies embedded into most LED light bulbs (especially the dimmable bulbs).

We recently purchased four new Shure QLXD wireless systems (in the G50 band) for our church to replace some old analog Mipro stuff that ran in the now outlawed 668 MHz band.  Shure's G50 band runs from 470 Mhz to 534 Mhz.  Immediately after installation of the new QLXD receivers into the audio rack at the mix position, I noticed 2 to 3 bars of RF signal with no wireless transmitters powered up!  After searching for every conceivable conventional source of such interference, I stumbled upon the cause when I came in one morning and did not turn on the lights before beginning my work (there was enough sun coming through the windows for me to see).  I noticed the RF noise was gone!  Was confused for a bit until I realized the only thing different this time was I had not turned on the lights.  No lights, no RF signal on the QLXD signal meters.  Hmmm..., so I turned on the lights, and presto!  The RF noise was back.  Did this four times to convince myself it really was the LED bulbs that were the source.

We have 10 hanging fixtures in the Sanctuary, and each fixture contains 12 bulbs.  There are eight A19 regular style bulbs and four BR30 flood style bulbs in each fixture.  The bulbs are sold by TCP Lighting.  I believe they are manufactured in China.  Here are the part numbers.

LED14E26BR3030K  (14-watt LED floodlight)
LED12E26A1930K    (12-watt LED bulb)

Some photos taken with my cell phone are attached to the end of this post showing the QLXD signal meters.  In a second post I will show some screenshots from Wireless Work Bench RF scans using one of the receivers.

[Bill Meeks]

Here is a screenshot from the Shure Wireless Work Bench software showing an RF scan taken with the LED lights in the Sanctuary on.  Notice the RF hash is everywhere across the band.  I also noticed that the intensity of the noise increased as the LED bulbs warmed up, and the peaks within the band also moved around as the bulbs ran longer.  After the bulbs have run for a while, the 534 MHz end is usually slightly quieter.  That's why I currently run my transmitters up there, but even then I have almost zero headroom in terms of RF signal between my transmitters and the noise floor.

[Bill Meeks]

Here is a WWB RF scan with the LED lights off in the Sanctuary.  I was amazed at the magnitude of RF noise those bulbs are generating all the way up at 534 MHz.  It just blew my socks off.

I did find that some clamp-on ferrite chokes reduce the noise substantially if the chokes are clamped onto the power leads directly at the bulb socket.  Because of the high frequencies involved, only the more expensive Mix 61 variety of ferrite works.  The cheaper and more common Mix 31 does not have enough impedance at 500 MHz to do much good.

In our case, I can only get to the power leads for the four BR30 bulbs in each fixture,  The power leads for the other eight A19 bulbs are not easily accessible, so choking the noise out with ferrite beads is not a good solution.  Looks like I will be hunting for new LED bulbs that are much quieter with RF emissions.

[Keith Broughton]

Good grief!!

[Henry Cohen]

As the aforementioned Google search will show, this issue is well known among the LMR (land mobile radio) and cellular folk. The RFI is worse at VHF frequencies and the LED replacements for fluorescent fixtures are particularly bad. Many in fact do not meet FCC specifications under Part 15 for spurious emissions.

There are a couple of manufacturers of LED bulbs that have responded to this issue and now produce very RF quiet [and Part 15 compliant] products, even when used in significant numbers in close quarters. They do of course cost a significantly more than the typical Home Depot / Lowes offerings. IIRC, CREE has some commercial models that are pretty quiet (likely due to the fact they own an RF products division).

[Ade Stuart]

MK Dons Arena, UK has LED working lights that completely wipe out the 2.4GHz band, WiFi, HME coms, W-DMX 

[Bill Meeks]

As the aforementioned Google search will show, this issue is well known among the LMR (land mobile radio) and cellular folk. The RFI is worse at VHF frequencies and the LED replacements for fluorescent fixtures are particularly bad. Many in fact do not meet FCC specifications under Part 15 for spurious emissions.

There are a couple of manufacturers of LED bulbs that have responded to this issue and now produce very RF quiet [and Part 15 compliant] products, even when used in significant numbers in close quarters. They do of course cost a significantly more than the typical Home Depot / Lowes offerings. IIRC, CREE has some commercial models that are pretty quiet (likely due to the fact they own an RF products division).

Thanks Henry.  Cree is one of the manufacturers on my short list for testing.  I have seen some Google posts from others praising them for being of higher quality with lower RF emissions.

It's a shame that cheap drives the world these days.  In the bulk amounts these manufacturers buy stuff, I would think it might cost some fraction of a dollar extra per bulb to include an RF choke and filter capacitor on the power supply board in the bulb.

Bill

[Henry Cohen]

In the bulk amounts these manufacturers buy stuff, I would think it might cost some fraction of a dollar extra per bulb to include an RF choke and filter capacitor on the power supply board in the bulb.

It's more than just a capacitor and choke. It starts with the PCB design and material, next the type and quality of the circuitry components, then proper shielding and testing. All in all, proper design, manufacture and compliance testing for commercial locations can more than double the cost of the final product.

[Don Boomer]

Just a side note ... if you search google you will find a lot of folks complaining their garage door openers no longer work after installing LED bulbs in their garages.

[Bill Meeks]

Just a final follow-up on this issue in the event it may help someone else in the future.  Replaced all the TCP Lighting LED bulbs with Cree equivalents and the RF noise is totally gone.  Signal meters on the QLXD receivers now all show zero bars with no wireless mic transmitters on.  With transmitters on, we get a full four bars of signal and no drop outs.

So the moral of the story is don't cheap out on LED bulbs if you intend to use them, and test a sample of what you intend to buy before you get the entire lot you may need.  In our case I purchased the minimum quanity I could of the A19 and BR30 Cree bulbs and tested.  Even held directly against the half-wave antenna while powered on in a lamp test fixture they produced no noise.  I swapped out the 120 LED bulbs for the new Cree models this past Saturday and our problem was solved.

[Keith Broughton]

Just a final follow-up on this issue in the event it may help someone else in the future.  Replaced all the TCP Lighting LED bulbs with Cree equivalents and the RF noise is totally gone.  Signal meters on the QLXD receivers now all show zero bars with no wireless mic transmitters on.  With transmitters on, we get a full four bars of signal and no drop outs.

So the moral of the story is don't cheap out on LED bulbs if you intend to use them, and test a sample of what you intend to buy before you get the entire lot you may need.  In our case I purchased the minimum quanity I could of the A19 and BR30 Cree bulbs and tested.  Even held directly against the half-wave antenna while powered on in a lamp test fixture they produced no noise.  I swapped out the 120 LED bulbs for the new Cree models this past Saturday and our problem was solved.

Thanks for posting a follow up.
Glad you got it sorted out

[Terry Martin]

That’s good info.  I’ve wondered about some of the cheaper LEDs affecting our IEMs.  We get noise occasionally - depending on the club we are at (same town, in some cases). 

[Rick Earl]

Just a final follow-up on this issue in the event it may help someone else in the future.  Replaced all the TCP Lighting LED bulbs with Cree equivalents and the RF noise is totally gone.  Signal meters on the QLXD receivers now all show zero bars with no wireless mic transmitters on.  With transmitters on, we get a full four bars of signal and no drop outs.

So the moral of the story is don't cheap out on LED bulbs if you intend to use them, and test a sample of what you intend to buy before you get the entire lot you may need.  In our case I purchased the minimum quanity I could of the A19 and BR30 Cree bulbs and tested.  Even held directly against the half-wave antenna while powered on in a lamp test fixture they produced no noise.  I swapped out the 120 LED bulbs for the new Cree models this past Saturday and our problem was solved.

I was lecturing to a college class about best practices for wireless mics, and I thought this would be a good example.   I went out to my local big box home store and bought the store brand LED bulb hoping to get some nice noise to show up on the analyzer.  I got it home and plugged in the light.... Nothing... I move it closer... still no difference in the noise floor.  Discouraged,  I started playing with other things around the house, I remembered an evaluation video lighting panel that I was testing for it's light quality, and  when I fired it up about 5 feet from the antenna the noise floor jumped almost 20dB!  This is a fixture that is supposed to retail over $1000.00.  It did drive the point home with the class.  It also was a great way to demo directional  antennas.   

[John Roberts {JR}]

I don't know that this is a diode "cost" issue but diode speed and a subtle noise interference related to reverse recovery current and carrier depletion. A common remedy in sensitive designs is to place a small snubber capacitor in parallel with the diode junction to damp high voltage rate of change. In several decades of general product design I never had a diode noise issue, but have seen and heard about the common mitigation strategy (capacitor snubbers).

I can imagine a low cost lamp manufacturer being reluctant to add JIC parts (snubber caps). They may be in violation of some FCC emissions regulations (not my area of expertise). So not a component cost issue per se, but clearly a less respectable manufacturer (if it wasn't tested for emissions).

JR

[Keith Broughton]

JR...I like your signature.
Had not noticed it before.

I guess there is no easy way to know if an LED light has been tested for RF noise unless specifically stated on the packaging.

[Rick Earl]

I don't know that this is a diode "cost" issue but diode speed and a subtle noise interference related to reverse recovery current and carrier depletion. A common remedy in sensitive designs is to place a small snubber capacitor in parallel with the diode junction to damp high voltage rate of change. In several decades of general product design I never had a diode noise issue, but have seen and heard about the common mitigation strategy (capacitor snubbers).

I can imagine a low cost lamp manufacturer being reluctant to add JIC parts (snubber caps). They may be in violation of some FCC emissions regulations (not my area of expertise). So not a component cost issue per se, but clearly a less respectable manufacturer (if it wasn't tested for emissions).

JR

This fixture is supposed to be "flicker free" for film and video applications,  It makes me wonder if it's the diodes or if there is some high frequency switching in the driver that is causing the noise. 

[John Roberts {JR}]

This fixture is supposed to be "flicker free" for film and video applications,  It makes me wonder if it's the diodes or if there is some high frequency switching in the driver that is causing the noise.

LEDs are DC devices (only conduct in one direction). To make light from AC the voltage must be rectified.

Driving LEDs from half wave rectified 60Hz will generally not be perceived as flicker, but full wave rectification bangs them 2x per cycle (Hertz) for even less on/off duty cycle.

JR

PS: Is the subject lamp FCC listed?

[Bill Meeks]

The LED bulbs that caused our problem were in the neighborhood of 4 to 5 years old.  They were bought and installed before I was responsible for the sound system.  The stamp you want to see on LED bulbs is FCC Part 15 compliant.

LED bulbs that are made to screw into the old standard 120 VAC sockets contain very small switching power supplies on a little circuit board in the bulb's body.  The power supply is converting the AC from the socket into a DC voltage for the LED modules that give off the light.  The RF noise comes from poor design of that switching power supply.  The switching supply is used to gain efficiency and dispense with the need for a large conventional iron-core transformer, and it also comes into play with the dimmable bulbs (which our troublesome ones were).

I think all of the bulb manufacturers have been getting better in terms of RF noise output, so it may be that most bulbs in the retail stores today are OK.  I would say that's especially true if they are made by Cree.  Cree also makes some "house brands" I believe for the big-box stores.

Bill

[Keith Broughton]

  The stamp you want to see on LED bulbs is FCC Part 15 compliant.


Bill

Just checked some bulbs I bought and , in very small print, it says "this device complies with part 15 of FCC rules"
Good to know.

[Rick Earl]

PS: Is the subject lamp FCC listed?

It is a beta test product, I was asked to evaluate it for it's light quality, it just happened to be handy when I was looking for UN-intentional transmitters for a wireless demonstration.  I let the manufacturer know about the RF noise, they said they were aware and working on some changes to address the issue.

[ProSoundWeb Community]