[RFI] Yet another balun question

[Jim Brown]

On Wed,6/29/2016 1:43 PM, Hare, Ed W1RFI wrote:
> >From what I know, a common-mode choke can control common-mode currents in a combination of two ways. The most effective is to absorb the unwanted signal and dissipate it as heat in the core.

Hi Ed,

The chokes that I recommend use this method. The core is selected to be 
very lossy in the frequency range where suppression is needed, and the 
number of turns carefully puts the low Q resonance in that range. The 
choking Z is the impedance at resonance, and the higher the choking Z, 
the more common mode voltage it can suppress.

We can model a choke in NEC using its equivalent circuit (which we must 
determine by measurement) by adding it as a load at the point we connect 
it. We can predict dissipation in the choke by setting NEC to use the 
actual TX power, and NEC will give us current in the choke and the 
dissipation. For the simple example of a dipole, we would model the coax 
as a single conductor going to ground from where the shield is 
connected, and put the load (the parallel resonant circuit) in the 
segment nearest the feedpoint.  Since the coax is NOT a transmission 
line in the common mode circuit, we would model it as a wire having a 
diameter equal to the coax shield and with insulation equal to the outer 
jacket. In other words, Vf roughly 0.97.

These chokes do NOT see the differential mode field, EXCEPT as Dean 
Straw, N6BV, described in his excellent QST piece a couple of years ago. 
That is, the length of line wound around the core sees the I squared R 
loss of the differential signal. This can be quite small if the line is 
matched, but can be quite large if it happens to be at a current maxima 
in very mismatched system (which is established by the feedpoint Z of 
the antenna and the electrical length of the line between the choke and 
the antenna.

> It can also be reactive, presenting an impedance that returns the current back down the line, to its source, where some of it will be lost as heat in the inefficiencies of the source, some lost as the signal getting into the source is propagated back onto the AC mains or to ground and some of it radiated along the way back, presumably farther away from the victim.  In any case where the victim does not rely on a signal received on an antenna, keeping the common-mode signal from getting to the victim by any means will reduce the potential of overload causing interference.

Here's how I think about what I think you're describing.

When a common mode choke is mostly reactive and with low resistance 
(like a coil of coax with no ferrite core, or a coil of coax on a low 
loss core, like #61 or #67), it will interact with the length of the 
line as an antenna. For example, a feedline that's shorter than a 
quarter wave (Vf = 0.97) will look capacitive, so a choke that's 
inductive will reduce the common mode Z, so the common current will 
increase. Worst case, it forms a series resonance.

OTOH, if the line is inductive, a low loss inductive choke WILL be 
effective. The key to understanding is realizing that the common mode 
circuit is an antenna, NOT a transmission line.

73, Jim K9YC


>
> -----Original Message-----
> From: RFI [mailto:rfi-bounces at contesting.com] On Behalf Of Jim Brown
> Sent: Tuesday, June 28, 2016 6:29 PM
> To: rfi at contesting.com
> Subject: Re: [RFI] Yet another balun question
>
> Hi Paul,
>
> First, let's call a spade a spade -- it's not a "balun" or even a "current  balun," it's a common mode choke.  What we CALL it helps us understand what it is and what it does.
>
> See more comments below.
>
> On Tue,6/28/2016 1:47 PM, N1BUG wrote:
>> I am sorry to ask this, but the probably obvious answer isn't obvious
>> to me.
>>
>> Assume I want to make a current balun consisting of two (or more)
>> chokes in series. Assume each choke will consist of a single 2.4"
>> toroid core wound with RG-303 coax (RG-58). Question: What should I be
>> looking at for physical orientation of one choke to another, and
>> distance between them? I am assuming that if nothing else, capacitance
>> between turns on one choke and turns on another choke would be a
>> consideration. Any guidance?
> Orientation of one choke to another is not critical. For all practical purposes, the field is confined to the ferrite core .
>
>> Second question: I think I read somewhere that there is little
>> difference in performance between all turns on a single such choke
>> wound "sequentially" around the core vs, winding half, then passing
>> the coax through and 180 degrees across the core, then winding the
>> remaining turns such that the coax leading into the choke and the coax
>> leading out can be 180 degrees opposed to each other.
> As far as I know, this method was proposed by W1JR for a choke he wound on a #61 core. I've wound such a choke and measured it, and I've never been able to see a difference between it and a sequential choke. Joe is a fine engineer, but I've shown (in my tutorial) why #61 is a terrible material for a common mode choke for the HF bands.
>
> You didn't say how much power you want to run through it or what the antenna is.  For 80M to about 15M, one of the bifilar chokes I've described wound with 12 turns of a pair of #12 enameled wire or #12 THHN should provide about 5K ohms of choking Z. If the antenna is reasonably well balanced and resonant, it should be good for at least 600W. For higher power and/or significant imbalance, two such chokes in series would be a good solution. With enameled wire, Zo is about 50 ohms. With THHN, it's closer to 80-90 ohms. Jerry Sevick described chokes like this in one of the later versions of his classic work, and noted these values for Zo. I've made both types and measured them, and got the same result.
> This is for closely spaced pairs.
>
> 73, Jim K9YC
> _______________________________________________
> RFI mailing list
> RFI at contesting.com
> http://lists.contesting.com/mailman/listinfo/rfi
>