Low EMI lighting for labs and work areas

[HalFoster]

All:

I'm in the process of moving and am thinking about the design of my new lab/work area and am trying to decide on the lighting.  In particular I would like to minimize the effect of radiated noise from the lights but also other factors such as the layout, CRI and color temperature.  Does anyone have any recommendations or experience they would share?  Thoughts so far are to have 4400K in general and 4800-5000K lighting over the task areas.  Having a high CRI is a plus but not strictly necessary.  The lowest noise solution would be custom LED arrays with a linear power supply - fairly straightforward to do but expensive.  Currently I am using 54W HO T-5 fluorescent fixtures in 4400K but they are fairly noisy in emissions so I would love to find alternatives (preferably COTS.)

TIA,

Hal

[Conrad Hoffman]

I don't have a good answer other than incandescent. Commercial CFLs and LEDs are all noisy. Whenever I  do noise work I have to replace my desk lamp LED with a regular bulb and sometimes turn off the overhead LEDs too. If you have dimmers they have to be turned up all the way and/or have better filters than they usually do. Watch out for things like humidifiers and dehumidifiers because they use similar circuit to control the fan speed and rarely have enough filtration. I had one dehumidifier that actually had filter component locations on a PCB, but unstuffed. I populated the board and the RFI was greatly reduced. Sometimes you can use an old portable AM radio with a telescoping antenna to sniff for noise.

[cdev]

LEDs powered by DC don't have much RFI if any.

[T3sl4co1l]

My LEDs are quiet... but then, I built them (with requisite filtering and testing). 

Alas, this isn't anything you'll ever see advertised about products -- at best you can go look up the FCC filings (and similar) and see how they were tested, and to what level.  Note that most will cheat e.g. power supplies tested with a load resistor on a short cable.  Lighting systems I don't know about, if they can get away with that (most installs are going to have long cables and large LED strips/panels on them?), but I wouldn't be surprised; and if nothing else, the fact that it's only so much length of cable and strip and panel, means EMI at low frequencies may roll off much higher than the total output, i.e. still emitting substantial near field noise that you'll see on the scope.

Tim

[Kleinstein]

For the LED lamps there are different technologies. Some are relatively quite, using some linear current setting, while other use a SMPS to provide DC voltage to power the LEDs.

[SilverSolder]

I don't have a good answer other than incandescent. Commercial CFLs and LEDs are all noisy. Whenever I  do noise work I have to replace my desk lamp LED with a regular bulb and sometimes turn off the overhead LEDs too. If you have dimmers they have to be turned up all the way and/or have better filters than they usually do. Watch out for things like humidifiers and dehumidifiers because they use similar circuit to control the fan speed and rarely have enough filtration. I had one dehumidifier that actually had filter component locations on a PCB, but unstuffed. I populated the board and the RFI was greatly reduced. Sometimes you can use an old portable AM radio with a telescoping antenna to sniff for noise.

I have used a big loop of wire as a "search coil" on an oscilloscope to find EMI sources.  E.g. one area I worked in had the ceiling lights wired incorrectly (daisy chained fluorescents where the return line was far away from the chain, creating a huge loop area) - which caused EMI so strong that CRTs in the area showed images that were vibrating from the AC field! 

[T3sl4co1l]

Well that doesn't even need to be EMI, just a few mT at mains frequency will do it.

Tim

[SilverSolder]

Effectively a room-sized air core transformer, lol

[TimFox]

Normal (as opposed to carefully designed and filtered DC-supplied) LEDs and CFLs also have substantial conducted EMI into the mains, when you are doing low-noise stuff.
The regulations on EMI concentrate on interference with RF applications, including (but not limited to) communications systems.
For examples,  https://www.govinfo.gov/content/pkg/GOVPUB-C13-1f770f754c83c95bcd5d338e84d0cb76/pdf/GOVPUB-C13-1f770f754c83c95bcd5d338e84d0cb76.pdf

[Ian.M]

Fluorescent tubes with 'old-skool' magnetic ballasts and glow starters are generally pretty low EMI once they've 'struck'.   Magnetic ballasts are still available new from specialist suppliers and can frequently be found in salvaged fittings.

[r6502]

Hello all,

I'm using LED stripes, assembled to an AL Profile. I can connect then to my ZigBee PWM control, or if I do not want to disturb measurements, I can connect them to a linear regulated power supply, here I use a standard lab supply in CC mode with Vmax set to max operating voltage of the LED stripe - works perfect. As connector I just use a standard DC connector 5.5mm outside, 2.5mm inside.

Guido