3
\$\begingroup\$

Almost every LED light I've tested (hold an AM or shortwave radio antenna nearby) emits RFI or EMI across a vast frequency range. The EM noise comes either from the base of the bulb, or from some box or wall-wart supplying a DC voltage to the LED fixture.

Is there any common solution to fixing or reducing this EMI? Other than throwing all these LEDs away and going back to hot incandescents? (which may or may not be legal in some settings.). Putting a opaque Faraday case around a light bulb makes them useless for lighting.

Maybe some sort of filter boxes on the AC wiring, or the DC wiring if accessible? A box of ferrite cores?

Does there exist any certified power efficient low EMI lighting? (e.g. more than just generic required regulatory labels on the boxes which may or may not indicate anything about what's actually in the box).

\$\endgroup\$
3
  • 1
    \$\begingroup\$ What specs? CE? FCC? Tempest? LEDS can run on filtered DC with no EMI \$\endgroup\$ Feb 22, 2020 at 16:51
  • \$\begingroup\$ Purchased off the shelf from random big box retail and hardware stores in the U.S. Specs vary from small table lamps and clip lights to bulbs for large floor lamps and overheads and outdoors. All AC line connected, direct or via some box (supplied DC converter). \$\endgroup\$
    – hotpaw2
    Feb 22, 2020 at 17:34
  • \$\begingroup\$ Spec is that I would like to turn on some lights without wiping out shortwave DX reception across multiple bands. \$\endgroup\$
    – hotpaw2
    Feb 22, 2020 at 17:42

4 Answers 4

2
\$\begingroup\$

You need an EMI choke as close to the socket as possible. The wires supplying power to the lamp are acting as transmitting antennas and radiating RFI. This also could include the house wiring in the walls.

At shortwave frequencies, the lamp itself is too small in relation to a wavelength to radiate very far, unless your shortwave receiving antenna is close to the lamps.

\$\endgroup\$
1
\$\begingroup\$

View these LEDs as EMI emitters just like black-brick battery chargers are EMI emitters.

If the LED is fed unfiltered non-DC, with fast edges, then only the faraday cage will suffice.

And if the black-brick should be poorly shielded (to block outside views of internal 200volt/100nanosecond edges), then that would need the faraday cage.

Otherwise, input shielding (coils, caps, ferrites) will improve the EMI.

Happy DXing.

\$\endgroup\$
0
\$\begingroup\$

My guess is STP cable or equiv. In a heavier guage will be EMI quiet with 12V on any earth grounded DC should not interfere with DX from a PC PSU.

\$\endgroup\$
0
\$\begingroup\$

Differential-mode emissions along the wiring is the most likely source.

Slug each line with an RF choke, as close to the source as possible: outputs if a wall wart, inputs if it's in the base. (No good doing one big choke round the whole cable, as this will not address differential-mode RF)

If the source's mounting has metal bits, earth them all to each other.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.