Topband: Elimination of Treadmill RFI on 160 meters

[Tom W8JI]

> Consider common mode noise on a simple parallel wire transmission line.
> Identical in-phase noise currents would flow on each of the parallel 
> wires.
> A common mode choke around the line would insert a high impedance equally
> onto both wires.The choke's effectiveness at suppressing the common mode
> currents would depend on the shunt impedance to ground of the two wires. 
> The
> shunt impedance between the two wires e.g., the impedance of the
> transmission line, is immaterial since there is no common mode voltage
> difference between them.
>
> On the other hand, the choke's ability to shield differential (as opposed 
> to
> common mode) currents depends a great deal on the differential shunt
> impedance. The lower the shunt impedance, the more effective the choke.
> Seems to me this is quite apparent if one draws out the circuit and 
> includes
> both the impedances to ground and the differential impedance between the 
> two
> wires. But, like I said, maybe I'm not following something.

Differential current suppression is a pretty complex function in a bifilar 
or trifilar choke, because mutual coupling between conductors enters the 
equation.

Bifilar winding of power leads is a two edged sword. While it reduces core 
flux from power line currents, it also reduces differential suppression in a 
similar fashion. In a tightly coupled choke, the windings simply behave like 
a transmission line.

It takes a stand-alone choke in each conductor to add significant 
differential mode impedances, or a poorly coupled common core choke. You 
will not obtain that with a bifilar or trifilar winding.

All of these problems, regardless of strong opinions, translate to ratios of 
shunt impedances and series impedances. They are nothing more than 
"pi-attenuators", or "pi-filters".

If impedances loading a choke on either or both ends are very high, 
attenuation is significantly reduced.  We can see this with choke baluns. If 
a choke balun is placed where the antenna sources very high common mode 
voltages (high impedance), choking impedance often has to reach impractical 
levels. The same is true for the load side of the choke.

As you say ZR, it is all about all of the impedances. It is a system, and 
looking at one part in isolation is pointless.

I keep bringing up a CATV system I had to cure for BCI. An apartment complex 
was constructed right over the radial system of a 5 kW AM station. The cable 
TV company tried triple shielded cables, and regular common mode chokes. 
They also tried CATV high pass filters. None of it worked, the system was 
mess.

All I did was have the special cables removed, and standard cables 
installed. All filters and chokes were removed. The cable entering each 
building entered alongside power mains, and the shield was bonded to the 
mains ground. The cable followed the power distribution to each apartment. 
Inside the apartment, the CATV wall plate was commoned with the safety 
ground of the wall outlet for the TV set though a suitable rating capacitor. 
Nearly every apartment cleaned up completely without using a single choke or 
filter.

This was because there was no differential RF allowed between the CATV and 
the mains. The common mode of the TV, since it was very high at 1520 kHz, 
prevented significant common mode to the set. A CM choke simply would not 
have done anything, because the TV's were already electrically small and had 
a pretty high CM impedance.

I have minimal CM choking here. I have no RFI issues, even beaming the four 
square through the house (just 75-150 feet away).  This is because I don't 
allow differential between wires and cables by use of outlet strips that tie 
the MATV antenna and other things to the power feeds at each hub. When I do 
need a choke, it isn't an astronomical value or difficult design. Just a 
snap on will often be enough, because the shunting impedances are kept low.

What works for ingress also works for egress.

73 Tom