[RFI] Linksys Routers, EMC, and Ferrite Chokes

[Jim Brown]

On Fri, 28 May 2004 12:01:36 -0400, Martin Ewing wrote:

>Fraser Robertson, G4BJM, kindly supplied some real Ethernet spectrum 
>measurements which I've posted at my  RFI web page -- 
>www.aa6e.net/aa6e/rfi .

I've been doing some research here in my office/ham shack and have learned
some interesting things.  I have two Linksys beasts in my office -- one of
the BEFR wireless routers that is cited in the class action lawsuit, and one
of their eight port switches. Both radiate lots of trash in the HF spectrum,
and I have identified them as the cause of strong carriers at 3511.25 kHz,
10106 kHz, 10122.56 kHz, 14031.27 kHz, and 21052 kHz.  

The HF radiation from these boxes has a very strong common mode component. I
say this because I am able to reduce their strength quite significantly (2-4
S-units) by winding 6-8 turns of the offending Ethernet cable around ferrite
toroids. The toroids I'm using are mostly Fair-Rite #61, but I've also used
some #43 cores.   

I've also learned something quite important about toroids that was clearly
noted in Fair Rite applications notes but not quantified. In general, the
peak attenuation frequencies of ferrites are moved down in frequency a LOT by
taking multiple turns around the ferrite. Thus, a #43 and a #61 material,
although generally considered to be effective only VHF and low UHF for
suppression, can be quite effective in the HF spectrum if you put enough
turns around them. A #43 core can provide a lot of impedance at 160 m, and a
#78 core will get you to the low end of the AM broadcast band!  

I'm off to the computer store now to look for an alternative to these Linksys
dogs. The cores to suppress the RF trash cost more than the boxes that
generate the trash!   

Jim Brown  K9YC