EMF Cancellation in Coax vs. Twin Speaker Wire

[pnielsen]

Can someone please help me with the following thought experiment?

There are straight two lengths of cable. One is medium gauge twin speaker ("figure 8") and the other is single core 75 ohm coax. Each length thus has two conductors. On each cable, I then separately short its distant end so current is flowing oppositely in each adjacent conductor. The near ends are connected individually to the same type of audio signal source.

In which cable, comprised of two conductors, will the EMF "cancellation" be more complete? What electrodynamic effect, if any, does the greater surface area of the screen have in the case of the coax?

Finally, would cancellation be more effective with higher voltage and lesser current, or vice versa? I assume here an even trade-off between current and voltage, e.g. 1V at 12A or 12V at 1 amp.

 

[dknguyen]

The greater surface area of the outer layer in the coaxial will have lower impedance to the high frequencies more currents will return via the conductors and currents energy will return capacitively through the air. (Currents that return capacitively manifest as radiated noise).

If you're just talking about cancellation of the EMF of DC currents, I'm not sure. I imagine the coaxial is better just because everything is more symmetric so the fields going both ways are more symmetric and on top of each other to cancel things out more.

 

[crutschow]

What is you audio application, that you are concerned about EMI (not EMF) cancellation?

 

[pnielsen]

I am interested in maximizing EMF cancellation (not EMI). There is no practical application at this stage.

dknguyen's second paragraph was along the type of information I am looking for. The currents are audio frequency and could be considered as DC.

There is a third possible wire configuration, apart form the side-by-side and coax I originally posted. That would be side-by-side but twisted, as they used to do in radio and hi-fi equipment.

Any further comments regarding the "cancellation" properties of these three configurations would be appreciated.

 

[crutschow]

The currents are audio frequency and could be considered as DC.

It makes no sense to me why you are concerned about EMF (ElectroMotiveForce) cancellation in a wire.
What EMF are you referring to?

 

[pnielsen]

EMF meaning electromagnetic field. I am referring to cancellation of energy vectors.

 

[rjenkinsgb]

EMF = Electro Motive Force; measured in voltage.
EMF is the equivalent of pressure in a fluid circuit, which is then measured or stated in the appropriate units.

Using confusing abbreviations makes the whole concept harder to understand...

 

[pnielsen]

My apologies but EMF is a widely used abbreviation for "electromagnetic field". In any event, I assumed my original post was sufficiently described to convey the intent of my inquiry. If necessary, I would be happy to clarify any remaining points so as to receive a specific answer. Thanks again.

 

[rjenkinsgb]

EMF is a widely used abbreviation for "electromagnetic field"

Never in any technical documentation!

Unwanted (non-deliberate) electromagnetic radiation or emissions are classed as "EMI" - Electromagnetic interference.
Note that any alternating magnetic or electric field results in some level of electromagnetic radiation.


Back to the original question:
There are two modes of emission from the cables.

Purely magnetic, which will cancel to some extent due to the opposite flow in the two conductors.

"Antenna" effect due to the applied voltage. The two conductors act as a voltage divider so the far end (shorted point) of the cable will have some voltage on it relative to the ground of the audio signal source (assuming it's a single ended 0v & audio type signal generator).

Radiated signal due to the antenna effect will cause a slight imbalance in current between the two conductors.

The coaxial cable should have the lowest EMI level as the symmetrical construction means the magnetic field has best balance.

The figure-of-eight cable, assuming it is untwisted, can be considered as a very long thing single-turn coil so the magnetic field around it will be much stronger; a rather higher EMI level.

If the audio source was balanced so the outputs were symmetrical about ground, that would reduce or eliminate the antenna effect.
The figure of eight cable would then be better as it's resistance should be equal in both legs, so the junction would have no net voltage.

Coax is likely to have unequal resistance so still have some voltage at the end junction.


If the figure of eight cable were twisted evenly over its length, the magnetic field would cancel out at each half twist so the overall emission would be minimal. That's why twisted-pair cables and balanced signals are used in data communications and audio.

 

[pnielsen]

If I understand correctly, a linear "antenna" of balanced and twisted figure eight cable would result in the greatest cancellation. Followed by the untwisted but balanced figure eight, and then the coax (even if balanced I presume).

In the case of the coax, could cancellation be improved by raising the voltage of the applied signal?

With regard to balancing, I would not connect the wires at one end of the antenna, and instead feed an identical signal into both ends so each travels oppositely to the other.

 

[pnielsen]

Another thought. Would it be practical to simply wire a small value resistor in series with the core of the coax so that its resistance matches that of the shield?

 

[rjenkinsgb]

Feeding both ends with an opposite balanced signal would cause the cable to act as an antenna and radiate.

With a balanced signal from one end, the coax loses out as the resistance of the two halves of the circuit are different (core vs screen) so the endpoint junction will have voltage on it due to the unbalanced resistive divider.
Higher voltage just means more emission.

 

[pnielsen]

Feeding both ends with an opposite balanced signal would cause the cable to act as an antenna and radiate.

I am not following this. What I was proposing in that instance is to create two parallel currents in adjacent wires of the cable but flowing in opposite directions. In my mind, that seems more efficient than simply shorting the distant end.

I understand the problem you describe with the coax. But if I drive the coax from both ends as above, the voltage difference would occur on the core side only, not a common point. Why then could I not insert a compensating series resistor at the connection to the signal source?

When you say "higher voltage means more emission" do you mean that for any given wattage input it is preferable to prioritize currentover voltage to achieve optimal EM field cancellation? Can you please explain why this is so?

 

[pnielsen]

This is the coax wiring described in the second paragraph

above.

 

[dknguyen]

ignore

 

[crutschow]

EMF is a widely used abbreviation for "electromagnetic field"

Okay. My mistake.
emf is the abreviation for electromotive force and EMF for electromagnetic field.

 

[crutschow]

would cancellation be more effective with higher voltage and lesser current, or vice versa? I assume here an even trade-off between current and voltage, e.g. 1V at 12A or 12V at 1 amp.

I believe the EMF is proportional to current, not voltage.
The only voltage of interest is that introduced by the characteristic impedance of the line, and the line has to be very long for that effect to be notices at audio frequencies.

In general coax is more effective to shield EMF than a wire pair.
Twisting a wire pair does reduce the EMF radiated.

 

[rjenkinsgb]

emf is the abreviation for electromotive force and EMF for electromagnetic field

Agreed, though it's the first time I've seen "EMF" used other than as electromotive force in over 40 years of working in electronics..

 

[rjenkinsgb]

When you say "higher voltage means more emission" do you mean that for any given wattage input it is preferable to prioritize currentover voltage to achieve optimal EM field cancellation? Can you please explain why this is so?

For any given load (resistance, or impedance with AC) current and voltage are proportional - you cannot change one without the other.

 

[pnielsen]

I believe the EMF is proportional to current, not voltage..

So you are saying that a coil carrying 100V at 1 amp produces less EMF than a coil carrying 100A at 1V?

In the context of my original post, can someone please comment on my diagram above showing the series termination resistor as a proposed method to equalize (balance) the resistances of the coax core and shield?