[RFI] RFI Digest, Vol 116, Issue 5

[Jim Brown]

On 9/7/2012 9:56 AM, jim feldman wrote:
> Remember it's the controller not the motor itself,
> although the motor may be acting as the radiator, along with the mains
> wiring.

Not quite -- the controller generates square waves which drive the 
motor, the pulses have a fast rise time and are thus rich in high order 
harmonics, and the wiring TO THE MOTOR radiates the trash. The wiring to 
the motor also creates a MAGNETIC field that is proportional in strength 
to its loop area, and RF bypass capacitors can combine with ground 
conductors to create large current loops.

Two important elements of the solution are 1) reducing that loop area by 
making the wiring that carries that current twisted pair; 2) slowing 
down the rise time of he pulses by adding a ferrite core to one 
conductor carrying that current.  However -- #2 should be done quite 
carefully, because dissipation in the switching transistor(s) increases 
as rise time gets slower.  But the strength of high order harmonics is 
proportional to rise time, so a little bit could go a long way.

> Try and keep filters as close to the controller as possible.
>   Check to make sure the chassis ground is still tight.  More turns around
> the same core are way better than more cores.  Different core materials
> tend to be better at different parts of the spectrum.

A serious common mode choke on the AC line is certainly a good move, and 
will kill whatever part of the noise is being radiated by the AC line.  
#43 and #31 materials made by Fair-Rite are all you need for the HF 
bands. Follow the winding guidelines in the link below.

http://audiosystemsgroup.com/RFI-Ham.pdf

73, Jim K9YC
>
> It's unlikely one thing will fix, so be prepared for multiple passes.