Dave,
I can't resist replying again, and challenging some of your points.
Coaxially wound CMC are not true common mode chokes.
I differ on that point. I claim that they are true CMCs, just like bifiliarly wound ones. They do suppress common mode current on a coax cable just as well as on a bifiliar one. Thus they tend to force the currents on both conductors to be of the same amplitude and opposite phase, regardless of whether the two conductors are arranged side-by-side or one-around-the-other.
They offer a
reflective function (inductive) for the RF energy on the outer surface of
the coaxial cable: a large +jX.
I would like to simply do away with the distinction between currents on the outside or the inside of the outer conductor, because it's irrelevant. We cannot separate them at the ends. And anyway, whenever the operating frequency is low enough, and the outer conductor is thin enough so that it's thinner than the skin depth, there physically isn't any separate current, but just ONE SINGLE current in the outer conductor. And a CMC wound with coax cable can work both below and above the limit frequency above which separate currents can exist on the outside and inside of the outer conductor.
Furthermore the distinction is irrelevant for this reason: At high frequencies and with thick outer conductors, the current on the inside of the outer conductor automatically is equal and opposite to the current on the inner conductor, regardless of whether or not there is a CMC. The CMC will then suppress current on the outside. And at low frequencies in a coax cable having a thin outer conductor, there is just one current in the outer conductor, which can be different from the current on the inner conductor, but a CMC will force them to be the same. So, the end effect is that regardless of whether those separate currents can exist or not, the CMC works the same. We just don't need to care about where exactly the outer conductor's current flows.
As such, they function as a current
'balun' reflecting the energy with a bit of absorption as well. A true CMC
accepts CM energy (coaxial cable, for example) at one port and 'outputs' DM
energy (equal amplitude with opposite phases) at the opposite port.
I can't agree with that logic. I can't see how a CMC, by itself, can turn CM energy into DM energy. Anyway energy involves voltage and time in addition to current. Letting the time aside, which means disregarding the confusion between energy and power, to talk about CM energy we need some external conductor (usually earth) to complete the circuit. As soon as such an external conductor is involved, we can no longer talk about the CMC doing anything, because it's the whole circuit that's doing it.
It is a bilateral device.
Yes. Any CMC is.
A bifilar wound CMC accomplishes the function of a
coaxially wound choke PLUS ensuring the DM required balance in amplitude
and phase at the DM side of the choke. This is accomplished by a
'feedback' mechanism between the transmission line on the toroid and the
induced magnetic currents within the core.
I absolutely disagree on that. There is nothing in a CMC wound with bifiliar line that would force such balance. At least not in a "clean" CMC that works reasonably free from parasitic effects. If you connect one port of a balanced-wound CMC to a ground-referenced signal source, and the other end to two different load resistors to ground, then the ground-referenced voltages on those two resistors will NOT be equal.Just the currents will be the same.
And if you do the same exercise with a coax-wound CMC, you will get the same results. Except if you have such strong parasitics (such as stray capacitance) that they dominate over the desired effects. But in that case you have a badly designed CMC.
This 'feedback' mechanism which
functions in both directions ideally cancels core magnetic currents induced
by each conductor of the bifilar windings. A coaxial 'balun' or current
balun such as coaxial cable wound on the core does not offer this
additional benefit of a bifilar wound CMC.
I disagree! The truth is that DM current causes no flux in the core, while CM current does cause flux, and that this happens regardless of whether the two conductors are side-by-side or coaxially arranged. For that reason the winding impedance of a CMC opposes the flow of CM current, while having no effect on the DM current. Again without any distinction between coax and parallel line.
The true CMC works on the same principle as parallel wire transmission
line, but without any added and lumped magnetic material. The interaction
between the two conductors of the oscillating RF field consisting of both
electric and magnetic fields on the line cancel eachother, resulting in no
radiation from the transmission line but only transmission of the RF energy
along the line. That's the physics (without the math) of the workings of
a true CMC. The presence of the magnetic material - the toroid -
'concentrates' the magnetic field produced by the bifilar windings much
like a dielectric 'concentrates' the electric field (in the case of a
capacitor), both of which allow for application of a lumped circuit
function instead of a distributed circuit function.
I believe that what you are trying to point out here is that with parallel wire the core does have some effect on the transmission effects along the line, while with coax cable it doesn't, given that parallel line has much of its field outside around the conductors, while coax line has it all confined inside. That's true. But the effect of this is mainly that the core will affect the impedance of the parallel line, if it's wound with the conductors very close to the core, while a coax cable is immune to this. It doesn't cause any effect on the fundamental action as a CMC.
CMC using parallel line should be wound with a reasonable spacing between the conductors and the core, to avoid increasing the transmission loss, and changing the line's impedance. Coax cable instead can be tight-wound around the core without a loss penalty nor a change in impedance. Of course only as long as its bending radius doesn't end up too small, but that's yet another effect that has nothing to do with what we are discussing here...
So I maintain that a winding of coax cable on a magnetic core makes a fully valid CMC, that even has some advantages over a CMC wound with parallel line. I don't accept that using parallel line to wind a CMC fundamentally produces any additional balancing effect.
Manfred
