[RFI] likely frequency for LED bulbs?

[Tom Thompson]

I just made some light bulb measurements using an RF current probe, an 
oscilloscope, and a spectrum analyzer.  My current probe is flat within 
3 dB from 500 kHz to over 50 MHz.  The common mode current on all bulbs 
was negligible through the entire HF band.  The compact fluorescent and 
LED lamps had energy below 1 MHz but above 1 MHz the energy was down 70 
dB.  The LED lamp had a switching speed of about 25 kHz.  There appeared 
to be no spectral components at the -90 dBm level between 1.8 and 30 
MHz.  The bulbs used were a 200 watt incandescent, a GE 26 watt helical 
compact fluorescent made in China, and a FEIT 60 watt replacement LED 
that uses 9.5 watts made in Taiwan and available at COSTCO.  I can email 
the spectral plots to anyone interested.

73,   Tom   W0IVJ

On 1/10/2015 5:52 AM, Greg Troxel wrote:
> wa4tuk <wa4tuk-rf at comcast.net> writes:
>
>> Thank you for posting your results. Information about what makes noise
>> and what is (relatively) quiet is hard to find. I will say I am
>> surprised. So far I have avoided LED lamps (A19 style) and use a mix
>> of incandescent and CFL where appropriate.
> I was surprised when I got the first one.
>
>> I would have thought the switching rate would have been in the 100kHz
>> range with harmonics radiating.
> Please note that I said I have other RFI sources; an attic antenna is
> not a good situation to start with.  I had dealt with a number of them,
> but the base level was higher than I expected yesterday.  So it's
> possible that these are problematic in quieter situations.  Also, I
> didn't sweep everywhere, so if there is a square wave spectrum that's
> pretty limited to harmonics, I could well have missed it.
>
> Given that these cost only $5ish these days (and that you may want to
> preserve your stockpile of pre-ban bulbs :-) it seems worth buying one
> or two to try out to see if you can hear them.
>
> (Power supplies for LED strip lighting seem to be an entirely different
> situation.)
>
> 73 de n1dam
> _______________________________________________
> RFI mailing list
> RFI at contesting.com
> http://lists.contesting.com/mailman/listinfo/rfi
>
>