Inductive Crosstalk Induced by Cable Braiding: Empirical Validation of Series Induced EMF

[SSS]

For me the inductive crosstalk is meaningless. With music I think it will not be perceivable. And last not least, just use two separated cables for the earphones.

 

[Redacted]

But we should be scared. Very scared.
FOMO kills!

 

[Holmz]

...
Unfortunately, the local frequency distribution characteristics of crosstalk are difficult to calculate through physical modeling. But the good news is that we can also use multi-tone measurement to help us understand the frequency components of the crosstalk signal, thereby adjusting the structure of the wire and changing the "crossfeed" coloration characteristics.

It is my understanding that in an IC it is capacitance that should be kept low.
And in a speaker cable is inductance that should be kept low.

I am pretty sure that inductance can be modelled.
Surely such physical modelling could be achievable?

 

[Winter_Ignition]

It is my understanding that in an IC it is capacitance that should be kept low.
And in a speaker cable is inductance that should be kept low.

I am pretty sure that inductance can be modelled.
Surely such physical modelling could be achievable?

I have conducted some new high signal-to-noise ratio tests, which show that this mechanism can indeed be modeled. And I would like to add here that not only in the frequency domain, but also in the time domain, electromagnetic induction crosstalk can cause a phase change of +90 degrees. I have already implemented the software algorithm simulation of this mechanism using python, merely discussing the mechanism itself without delving into topics related to timbre coloring. Moreover, inspired by the comment section, ground crosstalk does indeed play a dominant role in the low frequencies and is real. By using a stronger excitation signal, high-intensity mid-low frequency crosstalk outside the mutual inductance crosstalk mechanism has emerged.

The upper part of the picture shows the full-link crosstalk test of the headphones, which uses a stronger excitation signal. The lower part of the picture shows the comparison between the crosstalk and the original signal, after the frequency response characteristic is removed. It can be seen that there are obvious "modelable" patterns. However, I'm currently testing a limited number of samples, so I'm not sure if I can reach the same conclusion on all weaving lines.

 

[solderdude]

-58 dB cross-talk is inaudible in any stereo recording.
What is measured on the left side of your test fixture when only the right side is exited and the left side is covered using a sound damper ?

 

[Holmz]

-58 dB cross-talk is inaudible in any stereo recording.
What is measured on the left side of your test fixture when only the right side is exited and the left side is covered using a sound damper ?

Sure and one’s head, without cans on, can only get to 30-40 dB of channel separation.
But that is sort of the objective side and not really related to the physics in the cables.

I have conducted some new high signal-to-noise ratio tests, which show that this mechanism can indeed be modeled. And I would like to add here that not only in the frequency domain, but also in the time domain, electromagnetic induction crosstalk can cause a phase change of +90 degrees. I have already implemented the software algorithm simulation of this mechanism using python, merely discussing the mechanism itself without delving into topics related to timbre coloring. Moreover, inspired by the comment section, ground crosstalk does indeed play a dominant role in the low frequencies and is real. By using a stronger excitation signal, high-intensity mid-low frequency crosstalk outside the mutual inductance crosstalk mechanism has emerged.

The upper part of the picture shows the full-link crosstalk test of the headphones, which uses a stronger excitation signal. The lower part of the picture shows the comparison between the crosstalk and the original signal, after the frequency response characteristic is removed. It can be seen that there are obvious "modelable" patterns. However, I'm currently testing a limited number of samples, so I'm not sure if I can reach the same conclusion on all weaving lines.
View attachment 532668

I am not good at reading the Mandarin or Cantonese characters in the plot labels.

 

[EJ3]

The same headset cable, when measured on both the single and balanced ports, exhibits identical phenomena.
Ironically, most balanced headphone cables are coiled or braided, and their anti-interference properties are not effective against the cross-talk conditions I measured.

This looks like a + for my 2 pair of wireless headsets (both on the cordless phone frequencies [Sennheiser RS180 and Sennheiser RS220]).
I do not like being tethered (but I do like riding my lawnmower and listening to my music broadcasting from the house while wearing my headphones [30 meter range]).
The rest of the time, I listen to my speakers.

 

[Ropeburn]

Very nice work. And useful:

My conclusion is, the effect is more than small enough to not matter in practice at all. -60dB crosstalk very high up where most adults hear nothing, that's nice to know.

Physical, practical crosstalk might be higher than that. Non-perfect headphone/earpiece isolation, meaning stray sound escaping into the room and reaching the other ear, skull conductivity, these things combined must be way louder. Hehe!

Always nice seeing meticulous work done in this fashion, so we know exactly why it doesn't matter. Always smarter afterwards than before, so thanks for this.

 

[Holmz]

Very nice work. And useful:

My conclusion is, the effect is more than small enough to not matter in practice at all. -60dB crosstalk very high up where most adults hear nothing, that's nice to know.

Physical, practical crosstalk might be higher than that. Non-perfect headphone/earpiece isolation, meaning stray sound escaping into the room and reaching the other ear, skull conductivity, these things combined must be way louder. Hehe!

Always nice seeing meticulous work done in this fashion, so we know exactly why it doesn't matter. Always smarter afterwards than before, so thanks for this.

I think it is about 30dB.
The sound does wrap around ones head without cans on, so it is lot like the left side cannot hear the right side speaker.