[RFI] likely frequency for LED bulbs?

[Tom Thompson]

Hi Greg,

I suggest you build a fixture with a bulb socket on one end and an AC 
plug on the other end so you can light the bulb.  Before you put the AC 
plug on the cord, wrap a toroid (eg FT117-43) full of #22 enamel wire 
and terminate it in a BNC connector.  Put the AC cord through the toroid 
with 1 or 2 turns and then put on the AC plug. You now have an RF 
current probe.  Connect your radio (make sure not to transmit) with a 
short piece of coax.  Start with an incandescent bulb as a reference and 
record the noise.  The noise level should be negligible if the toroid is 
not acting like an antenna.  If it is, put it in an enclosure.  Next 
screw in your test bulbs and see what the noise is on the various 
bands.  If you have a calibrated receiver, the current probe should be 
fairly flat over the HF band.

Hope this helps,

Tom   W0IVJ


On 1/9/2015 12:22 PM, Greg Troxel wrote:
> I have 2 LED bulbs (60W incandescent replacement) and have been unable
> to hear them, which is good.  Today I bought 3 more types (at the local
> Ace, because I can return them if they make noise).  So I'm about to
> test them.
>
> I am curious what people's experiences are in terms of which frequencies
> are likely to be the problems.  I would sort of expect 40 meters, just
> because that seems to be the epicenter of RFI generally.   But I realize
> it's highly dependent on the switch-mode supplies in the bulbs.  I also
> saw a report of interference from LED bulbs meant to replace 60W metal
> halide bulbs.
>
> To do a quick test, my procedure is to put one of the new bulbs in a
> basic desk lamp in the shack, turn off other known noise sources, use an
> attic multiband dipole (one end of which is only 15' away), and step
> through the bands 160m-6m and see if I can tell a difference turning the
> light on and off.  I will also check 145 MHz.
>
> Does that make sense, or is there anything else that's easy that I can
> do quickly?  (I don't have a spectrum analyzer with a LISN....)
>
> I'll post my notes on specific brands and findings.
>
> 73 de n1dam
> _______________________________________________
> RFI mailing list
> RFI at contesting.com
> http://lists.contesting.com/mailman/listinfo/rfi
>
>