[RFI] Line filters for VHF?

[Jim Brown]

On Wed,5/20/2015 3:01 PM, N1BUG wrote:
> On 05/20/2015 03:31 PM, David Robbins wrote:
>> I wonder if the 3 cores you used stopped most of the junk from
>> that one box, but you were still seeing junk coming in from the
>> rest of the building or outside??
>
> I don't think so, because when that one box is powered off, the junk 
> on its power cord and wiring in the building goes away.

Ferrite cores clamped on a cable can be a good fix for common mode 
current, but they won't help with differential mode current. It is well 
known that coupling from differential mode current can be reduced by 
twisting the conductors that carries that current.

Commercial line filters ARE differential filters, so should help with 
that if properly applied, and if their reactive components don't have 
too much stray reactance -- that is, stray C in the chokes, stray L in 
the capacitors. (What the power industry calls "common mode" is voltage 
between neutral and green.) Properly applied means zero length leads 
between the filter and the noise source, including a zero length bond 
between the shielding enclosure of the noisy device and the filter.

Here's a series of CorCom filters that looks promising.
http://www.te.com/usa-en/product-1-6609037-1.html#pdp-docs-features 
Click on the "see all documentation" link, then choose the top pdf, 
which includes graphs. You might also try calling them or using Live 
Chat to ask about VHF.

As to the effectiveness of the number of cores and their type -- for a 
single turn (that is, passing the cable through it once), the type you 
chose, #43, is ideal for for 2M. Winding turns would lower its 
resonance, and thus its effectiveness to around 6M.  Choking Z will be 
proportional to the total length of the ferrite material. This is going 
to be somewhat logarithmic -- that is, for 6 dB, the current must be 
reduced by half, which corresponds to doubling that Z. So -- assuming 
it's all common mode, it would take 3 more cores to reduce that common 
mode current by half.

Ideally, the choke should be as close as practical to the noisy device. 
Fair-Rite #61 would be the weapon of choice for 440 MHz.

73, Jim K9YC