Do LED bulbs cause interference on audio circuits?

[Beatlejuiced]

I am in the process of setting up an amateur recording studio in my workshop which has overhead fluorescent light fixtures. I am changing the fluorescent lights to incandescent fixtures due to the tendency of fluorescents to cause interference noise in audio circuits, which in my case will be both line-level and microphone-level.



I had not shopped for incandescent light bulbs for some time, so when I went to the home improvement store to buy both the fixtures and the light bulbs I was somewhat dismayed to discover that LED bulbs had almost completely supplanted incandescents on the store shelves. I asked the sales clerk about them and he said that each LED bulb has a transformer integrated in the bulb chassis. Having no experience with LEDs I don't know if they can cause noise in audio circuits, so I am hoping that some members with more knowledge than I have can clue me in.


For now, I have stocked up on some halogen incandescent bulbs, but for future reference I would like to know the story on LEDs. The sales clerk told me that the store will stock incandescents for the forseeable future, but having witnessed the rise and apparent fall of CFLs and the vagaries of the relevant governmental agencies I am without faith in such pronouncements.


So, what's the real story? Has anyone experienced LED-induced noise in their audio circuits, and is there any reason to expect that they could cause such interference?

 

[xx3stksm]

LEDs itself are free from electrical noise. A transformer(SMPS) in the bulb chassis has poor quality AFAIK, especially high power LEDs(more than 100W). The quality difference between products is also significant. If you want to have reliable performance, it's better to use a linear PS for LEDs. If you use LEDs without a transformer, many LEDs are available out there, color, shape, and power.

 

[wiseoldtech]

Indeed, run LED's from a linear regulated supply.


I installed under-cabinet lighting in the kitchen back in May.
4, 1 foot 5000K strips, built my own variable supply (LM317+ T03 booster) with dimming feature - 32-34.5V.
Cost me only about $140 all included.
And it's reliable because it's built by me.
360+ equivilant watts of light now, with only a 36 watt AC draw.
And told it will last 45 years at 3 hours a day use.
Can't beat it.

 

[lcsaszar]

The situation is more complicated. For preinstalled fixtures running off 110V (or 230V in EU) the only option is to put compatible bulbs in the sockets. These LED bulbs have some SMPS built inside. I think interference travels along the cabling, not over the air. Perhaps some mains filter or fully balanced audio circuitry could prevent interfering with audio.
I also stocked some halogen incandescent bulbs for the future, at least until LED technology evolves and they give acceptable illumination.

 

[sgrossklass]

First of all, if plain fluorescent lighting is giving you trouble, You Have A Problem(R). With (a) all-balanced cabling and (b) devices following AES48-2005 guidelines for XLR pin 1 wiring, you should have no issues. We're talking a garage, not a huge studio. I mean, I wouldn't exactly run the audio cabling next to power lines (which may not be according to code anyway), but still.

4, 1 foot 5000K strips, built my own variable supply (LM317+ T03 booster) with dimming feature - 32-34.5V.

Hint: Set up the regulator as a variable current source - that's what LEDs ultimately care about. Your dimming must be terribly nonlinear (well, non-logarithmic). I find it a bit surprising that this detail would have eluded you until now.

FWIW, I have two Amazon Basics branded 7 W E27 screw-in bulbs that have been working fine with no RF noise whatsoever (I did test with a portable shortwave receiver). Same for about a dozen more 8 W jobs, except one that went flaky on me shortly after installing. I thought color was a bit on the greenish side, but other than that no complaints - no flicker, no nothing. Mind you, I'm in 230V territory, so you won't be able to buy the same over the pond, but still. Apparently it is possible to make these guys with all-linear power supply circuitry.

The usual approach for quality LED drivers is using a switch-mode converter with active PFC followed by a switch-mode current driver (basically the switch-mode equivalent to the linear supply + LM317-ish reg used as a current source). The main advantage of this is flexibility in panel choice and high efficiency over a wide range of line voltages, where a linear power supply has to be sized pretty exactly to match minimum expected line voltage and panel voltage at maximum output current.

At the kind of voltages that larger LED panels tend to need (often in the 30s of V), linear supplies aren't necessarily all that inefficient, and you should be able to hit about 80% or not much less at least if you choose a good-quality transformer. Off-the-shelf LED drivers typically seem to hit 85-90% or so, which is clearly better, but not so much that you would consider it night and day, so the simple and rugged linear supply definitely isn't out immediately.

 

[Beatlejuiced]

Thanks for the replies. My recording gear is all "semi-pro" with the ability to use balanced cabling between units and I use a power conditioner between my gear and the mains, but please note that I have not yet moved my gear into the workshop. I chose to use flush-mounted incandescent fixtures because the present fluorescent fixtures are also flush-mounted on the ceiling with in-wall wiring direct to the mains on an in-wall switch.



Other reasons for the change to incandescent fixtures is the fact that my workshop is not climate controlled, so in the winter before I can heat the shop with electric space heaters the fluorescent lights do not illuminate until they are warmed, and the fact that I simply prefer the quality of incandescent illumination to that of fluoresecent.


Does anyone have experience using standard Edison-style screw-in LED bulbs in proximity to or on the same mains line as their audio equipment, and have you noticed any interference?

 

[DF96]

Mains-driven LED bulbs vary widely in their interference capability, and they seem to have a short catalogue life so even if someone told you to use/avoid Blogg's bulbs you cannot tell if the Blogg bulbs you buy will give the same behaviour as the ones he bought a short while ago. You just need to experiment yourself. When you find a bulb which works fine, buy enough to last you as long as you need.

 

[MAAC0]

My experience with Chinese lamps usually is they affect FM reception.

 

[Osvaldo de Banfield]

I had disassembled several of them (Mainly Baw), and I found usually several diodes blown and a step down (Buck) PWM regulator made in one IC and a couple of low value electrolytics (2.2 to 4.7µF @ 400V) for 220V supplies. The are very bad as illumination.

And here, the led lights used as street lightning are very bad quality: flicker at 1-10Hz depending on the failure type, most are kaput and they incandillate when directly seen or reflected in car glasses. This is thanks to our politics that buy thrash at prices of gold.

 

[wiseoldtech]

Hint: Set up the regulator as a variable current source - that's what LEDs ultimately care about. Your dimming must be terribly nonlinear (well, non-logarithmic). I find it a bit surprising that this detail would have eluded you until now.

The usual approach for quality LED drivers is using a switch-mode converter with active PFC followed by a switch-mode current driver (basically the switch-mode equivalent to the linear supply + LM317-ish reg used as a current source). The main advantage of this is flexibility in panel choice and high efficiency over a wide range of line voltages, where a linear power supply has to be sized pretty exactly to match minimum expected line voltage and panel voltage at maximum output current.

At the kind of voltages that larger LED panels tend to need (often in the 30s of V), linear supplies aren't necessarily all that inefficient, and you should be able to hit about 80% or not much less at least if you choose a good-quality transformer. Off-the-shelf LED drivers typically seem to hit 85-90% or so, which is clearly better, but not so much that you would consider it night and day, so the simple and rugged linear supply definitely isn't out immediately.

My analog supply works fine.
Dimming (via rotary pot) is smooth from a candle glow up to full brightness with pure filtered/regulated DC to the LED strips.
With a 1.5A transformer the supply barely gets warm, and is mounted inside one wall cabinet.

 

[ndcraig]

Thanks for the replies. My recording gear is all "semi-pro" with the ability to use balanced cabling between units and I use a power conditioner between my gear and the mains, but please note that I have not yet moved my gear into the workshop. I chose to use flush-mounted incandescent fixtures because the present fluorescent fixtures are also flush-mounted on the ceiling with in-wall wiring direct to the mains on an in-wall switch.



Other reasons for the change to incandescent fixtures is the fact that my workshop is not climate controlled, so in the winter before I can heat the shop with electric space heaters the fluorescent lights do not illuminate until they are warmed, and the fact that I simply prefer the quality of incandescent illumination to that of fluoresecent.


Does anyone have experience using standard Edison-style screw-in LED bulbs in proximity to or on the same mains line as their audio equipment, and have you noticed any interference?

I've used normal incandescent bulbs for years in the studio with no problems.
The issue was with dimming. Using standard triac dimming introduced buzz. Even on separate panel and circuits -it stopped the electrical interference, but the filaments themselves would faintly buzz from the chopped waveform -enough to be picked up by the better condenser microphones

Switching to variac dimming solved all of those issues. We got ours from a local theater that was upgrading their ancient side-stage lighting controls and got a cabinet of six superior electric 25A variacs that we had wired into the lighting circuits in a hallway alcove between control room and studios.
The pure sine-wave power from that system was entirely noise free and would dim transformer-based halogen lighting, older neon and even motors if they were the right type.

 

[Osvaldo de Banfield]

Linear PSU's do have lots of noise in the primary and secondary circuits, mainly in the 3° harmonic of the power line frequency, except you use choke input filtering.