Symmetrical audio cable over single-wire cable

[lister]

Most newer earphones and most headphones comes in a form of single stereo cable.
 
Just wanted to ask if anyone noticed the difference between Symmetrical cable versus single stereo cabling.  Recently, I tested many earphones and found that there are serious cross talk between the left and right channels.
 
The ones without cross talk problem are the ones with L/R independent wiring to the connector.  This is what Sennheiser refer to as the 'symmetrical cable'.
 
1) Has this issue ever been brought up?
2) Is this a design flaw or oversight (in the single cable)?

 

[lister]

Anyone with comments?
Anyone care about distortion created by the cabling?
 
The cable is a big part in the signal path.  You can have the perfect source, perfect Amp, perfect Headphones, and yet everything is ruined by the cable!

 

[Steve Eddy]

I'm not sure what you mean by "severe crosstalk." How exactly did you test for it?

Tonight I did a little testing myself.

I pulled off a little over 12 feet of Canare 4S6 star quad speaker cable, which has a nominal capacitance of 125 pF per foot which is quite high.

I wired the right channel to a 1/4" TRS plug and at the other end, I soldered a 60 ohm resistor across it to match the resistance of my Audeze LCD-2s. This was done so that the right channel of the cable would have the same voltage and current as it would if plugged into the LCD-2s, but without being connected to a driver so the right channel would be dead silent.

For the left channel, I connected a 4 pin mini XLR for plugging into the left channel of the LCD-2. At the 1/4" TRS end, I soldered the ground lead of the left channel to the jack's ground. For the positive lead, I first took a 2 ohm resistor and solderd one end to the jack's ground, and the other end to the positive lead of the cable's left channel.

This was to emulate the output impedance of the amp. A photo of the cable assembly is shown below.

I then plugged in the cable and started playing some music.

I had to turn the volume control of the amp nearly all the way up before I could really hear anything. Using a normal cable, with the volume control in the same position where I began hearing crosstalk, the music was quite a bit louder than I would normally listen. At normal listening levels, I had to strain a bit to hear any crosstalk. And I seriously doubt that tiny bit of crosstalk would ever be audible while actually listening to music through both channels.

And keep in mind this was with 12 feet of a rather highly capacitive cable (it's capacitance that's responsible for crosstalk in a cable like this which has a very low inductance).

As for distortion produced by the cable, forget about it. Unless something is broken or you have a bad solder connection, any cable is going to be far far more linear than any amp and far far FAR more linear than any headphone.





se

 

[dprimary]

Crosstalk generated in cable should be greater then 80dB down. The only time cabling crosstalk can cause problems in and long analog multi pair snake. Long as in 100 to 2500 feet even then you need to have microphone level signals and and line level signals which about about 60 dB difference in level.   
Maximum cross talk on my head seems to be best case 12dB. 18 with a finger in one ear.

 

[analogsurviver]

I'm not sure what you mean by "severe crosstalk." How exactly did you test for it?

Tonight I did a little testing myself.

I pulled off a little over 12 feet of Canare 4S6 star quad speaker cable, which has a nominal capacitance of 125 pF per foot which is quite high.

I wired the right channel to a 1/4" TRS plug and at the other end, I soldered a 60 ohm resistor across it to match the resistance of my Audeze LCD-2s. This was done so that the right channel of the cable would have the same voltage and current as it would if plugged into the LCD-2s, but without being connected to a driver so the right channel would be dead silent.

For the left channel, I connected a 4 pin mini XLR for plugging into the left channel of the LCD-2. At the 1/4" TRS end, I soldered the ground lead of the left channel to the jack's ground. For the positive lead, I first took a 2 ohm resistor and solderd one end to the jack's ground, and the other end to the positive lead of the cable's left channel.

This was to emulate the output impedance of the amp. A photo of the cable assembly is shown below.

I then plugged in the cable and started playing some music.

I had to turn the volume control of the amp nearly all the way up before I could really hear anything. Using a normal cable, with the volume control in the same position where I began hearing crosstalk, the music was quite a bit louder than I would normally listen. At normal listening levels, I had to strain a bit to hear any crosstalk. And I seriously doubt that tiny bit of crosstalk would ever be audible while actually listening to music through both channels.

And keep in mind this was with 12 feet of a rather highly capacitive cable (it's capacitance that's responsible for crosstalk in a cable like this which has a very low inductance).

As for distortion produced by the cable, forget about it. Unless something is broken or you have a bad solder connection, any cable is going to be far far more linear than any amp and far far FAR more linear than any headphone.





se

Although commendable, dynamic headphones are the least likely to suffer from crosstalk in cables. This was a nice attempt to demonstrate, once again, that cables do not matter.
 
Which is ANYTHING that can be said about electrostatics. Historically, the cables did get a bit better with electrostatics - but first models were THE models how NOT to connect electrostatics.
 

Jecklin Float Electrostatic - early model
 

 
Second incarnation of Jecklin Float Electrostatic
 
- now at least a flat cable a la Stax, not a 5 wire round one from the original model - which is as horrible way of connecting electrostatics as it gets - and shared with Sennheiser Unipolar 2000 and 2002 ( but Senns at least ARE electrically symmetrical). Also a new transformer - the original had tiny transformers inside, with THD in the bass going trough any roof due to the saturation at any reasonable drive level in the bass. In addition, inductively coupled - separation on the order of 20 or so dB. An early attempt at crosfeed ( in earspeakers that DO NOT NEED CROSFEED to begin with - they were historically the first can to dispense with this necessity - by design ) ?
Both first and second gen still use exit of cable from the left hand side of the headset - making the capacitance in the right channel higher, simply for the additional length of the cable across the "headband", making the electrical (capacitive)  load asymetrical.
 

 
Recently re-introduced Floats - by Quad Atelier of Germany. From pics I've seen is still not clear whether each transducer has now its cable connected directly and not across the "headband" as in both predecessors. They seem to have even reduced the size of the connector ( necessitating increase in capacitance ) - with drivers that require MUCH more healthy voltage than any Stax ever made. Oh dear...
 

 
Stax Lambda ( model this or that ) - or cable standard no commercially available electrostatic can has been able to improve upon so far.
For starters - it has MORE capacitance than the driver itself ( even the latest and greatest models ) - and there is crosstalk between channels as the frequency goes up - it is "nicely" capacitively coupled. Still MUCH better than these 5 (or 6)  wires bundled into round cable of the first gen - where capacitive coupling/crosstalk was at the best/worst possible.
 
Now - anyone - dare to say that cables DO NOT MATTER;  it only depends on the application.

 

[Steve Eddy]

Although commendable, dynamic headphones are the least likely to suffer from crosstalk in cables. This was a nice attempt to demonstrate, once again, that cables do not matter.

It was a specific response to a specific question about crosstalk. Not any sort of general treatise. If you want to address anything I said in the context in which it was said, go right ahead. Otherwise, all this is just one big straw man.

se

 

[lister]

I'm not sure what you mean by "severe crosstalk." How exactly did you test for it?

Tonight I did a little testing myself.

I pulled off a little over 12 feet of Canare 4S6 star quad speaker cable, which has a nominal capacitance of 125 pF per foot which is quite high.

I wired the right channel to a 1/4" TRS plug and at the other end, I soldered a 60 ohm resistor across it to match the resistance of my Audeze LCD-2s. This was done so that the right channel of the cable would have the same voltage and current as it would if plugged into the LCD-2s, but without being connected to a driver so the right channel would be dead silent.

For the left channel, I connected a 4 pin mini XLR for plugging into the left channel of the LCD-2. At the 1/4" TRS end, I soldered the ground lead of the left channel to the jack's ground. For the positive lead, I first took a 2 ohm resistor and solderd one end to the jack's ground, and the other end to the positive lead of the cable's left channel.

This was to emulate the output impedance of the amp. A photo of the cable assembly is shown below.

I then plugged in the cable and started playing some music.

I had to turn the volume control of the amp nearly all the way up before I could really hear anything. Using a normal cable, with the volume control in the same position where I began hearing crosstalk, the music was quite a bit louder than I would normally listen. At normal listening levels, I had to strain a bit to hear any crosstalk. And I seriously doubt that tiny bit of crosstalk would ever be audible while actually listening to music through both channels.

And keep in mind this was with 12 feet of a rather highly capacitive cable (it's capacitance that's responsible for crosstalk in a cable like this which has a very low inductance).

As for distortion produced by the cable, forget about it. Unless something is broken or you have a bad solder connection, any cable is going to be far far more linear than any amp and far far FAR more linear than any headphone.





se

 
Hi Steve,
 
Thanks for posting with your experiment.  I was trying to decode your experiment based on what you've said.  I've sketched a schematic below.  Please correct me if I'm wrong...

 
My response is that, you have the right idea.
But I never doubted the XLR setup.  Although I don't own any XLR type equipment.  From what I've read, it's a fully balanced setup, which should give excellent common mode rejection.  And crosstalk is unlikely.  The XLR is almost similar to the symmetrical cable that I was referring to, but only better.
 
Your setup is upwards of $1k.  So I am pretty sure the MFG gets it right.  And whatever findings you've got from your experiment, I'm sure the result is correct as expected.
 
I think my original thread post might have been misleading.  So please let me clarify my issue/concerns (in the next/following post)...

 

[Steve Eddy]

Hi Steve,

Thanks for posting with your experiment.  I was trying to decode your experiment based on what you've said.  I've sketched a schematic below.  Please correct me if I'm wrong...

Great illustration. Yes, it is correct.

My response is that, you have the right idea.
But I never doubted the XLR setup.  Although I don't own any XLR type equipment.  From what I've read, it's a fully balanced setup, which should give excellent common mode rejection.  And crosstalk is unlikely.  The XLR is almost similar to the symmetrical cable that I was referring to, but only better.

Ok, I see where the trouble is and why it was so misleading to call headphone amps "balanced" when referring to the headphone output.

With regard to the headphone output, the more appropriate term to use is "bridged." However the first amps with bridged outputs were misleadingly called "balanced" and the industry has been calling them that ever since. I have tried to get people to use the correct terminology for years now but it has been hopeless.

I understood you to be referring to crosstalk within the cable itself, which you're going to have even with a bridged amplifier as long as the left and right channels share the same cable. So I thought you were referring to using separate left and right cables, instead of a bridged output.

Ok. With a regular TRS plug/Jack, you have the ground contract resistance that is common to both left and right channels. So by ohms law, any current that flows through that resistance will create a voltage drop that will also be common to both channels, which will result in some amount of crosstalk.

In a bridged amplifier, which uses four contacts, two for each channel, you eliminate that common contact resistance and therefore won't get any crosstalk from that resistance.

So to test for the crosstalk due only to that common contact resistance would require a different test setup. Sorry for the confusion.

se

 

[lister]

Great illustration. Yes, it is correct.
Ok, I see where the trouble is and why it was so misleading to call headphone amps "balanced" when referring to the headphone output.

With regard to the headphone output, the more appropriate term to use is "bridged." However the first amps with bridged outputs were misleadingly called "balanced" and the industry has been calling them that ever since. I have tried to get people to use the correct terminology for years now but it has been hopeless.

I understood you to be referring to crosstalk within the cable itself, which you're going to have even with a bridged amplifier as long as the left and right channels share the same cable. So I thought you were referring to using separate left and right cables, instead of a bridged output.

Ok. With a regular TRS plug/Jack, you have the ground contract resistance that is common to both left and right channels. So by ohms law, any current that flows through that resistance will create a voltage drop that will also be common to both channels, which will result in some amount of crosstalk.

In a bridged amplifier, which uses four contacts, two for each channel, you eliminate that common contact resistance and therefore won't get any crosstalk from that resistance.

So to test for the crosstalk due only to that common contact resistance would require a different test setup. Sorry for the confusion.

se

 
Hi Steve,
 
Yes, you've got my original concern when stating 'shared the same cable'.
 
 
Dear all,
 
Please allow me to clarify my thread's original question/concern...
 
I was not questioning Hi-End/Pro audio equipment.  My original concern was on the amount of cross talk on the budget ear-buds & headphones, due to bad choice of cabling.  Now since budget is a relative term, I'll just put a ball park number of say ... below $300 USD.
 
To illustrate my point about the cabling and crosstalk, please see the two images attached here...

 
The two images posted are (1) Monster Turbine (2) Sennheiser MX370.
As you can see the Turbine bundles both L&R conductors into one cable, whereas the MX370 clearly have isolated conductors separated all the way from the driver to 3.5mm connector.
 
The test result (without going into details) is that the Turbine had bad L&R channel crosstalk/leakages.  And the MX370 did not have any crosstalk issue.
By the way, these are only examples that I have tested.  There are much more ear-buds and the so call 'IEMs' having this issue!
What irritated me was that the fact that the Turbine cost me $150 USD, and the MX370 was only $30 USD.  I just cannot believe a $30 earphones got it right, but the more expensive Turbine failed on this simple test.
 
Since then I've tested several other ear-buds, adapters, as well as extension/aux cables.  And most have minor to serious crosstalk issues.
And by the way, the Sennheiser HD650 3.5mm-to-1/4" adapter (pig-tail, mfg original) has crosstalk issues!  So try to avoid using that pig-tail, use a single piece adapter instead.
FiiO L16/L17, RC-MH1, and RC-UX1 doesn't seem to have this crosstalk issues.  They happen to be the cheapest that I could find with symmetrical cabling.
 
Anyway, I'm now really skeptical when checking out new earbuds and headphones.  Because mfg seems to have miss the mark on the cable design.
 
For those who'd like to try, it's really simple to test ear-buds for crosstalks.  NO need to cut/patch/solder anything.  Here are the steps...
(1) Use any audio editor (Audacity is free), create a stereo test tracks and generate a tone on the Left channel and zero on the Right (basically mute one side).
(2) Now, put on the Right ear-buds, but take away the Left and cover that with a pillow (so you don't hear sounds from the LEFT bud).
(3) Finally, play back the test track (at nominal listening level), and listen to small test tone appears on the Right bud.
(4) You can repeat these steps swapping Left and Right.
 
It's not an accurate test, and it won't tell you exactly how many dB is the crosstalk.  But I assure you that you'll hear the difference between a good one and a bad one.
 
If any of you tried, please share your findings.  I don't have all the IEMs out there on the Market, but I'd like to know which one has or has not the crosstalk problem.

 

[dprimary]

Are you sure you don't have a problem with a jack or the plug not seating all the way? I can't get enough acoustical isolation to tell if I can hear any electrical crosstalk. I can measure it at the office in the next day or two.

 

[lister]

  Are you sure you don't have a problem with a jack or the plug not seating all the way? I can't get enough acoustical isolation to tell if I can hear any electrical crosstalk. I can measure it at the office in the next day or two.

Pretty sure not a connection issue.  I usually like to cross check my work with various setups (is redundancy the right word?).  I've tried on Desktops, Laptops, as well as external DAC (Xonar Es one).  The Eone has 'bridged' XLR outputs, but I don't have XLR phones, so too bad.
 
By the way, you should be able to hear it in a typical office environment (~ 40dB).  If necessary, an industrial 3M ear plugs might help.
 
Also, we are not talking about 100 feet long cable.  Only a typical of 3ft for ear-buds/IEMs.

 

[Steve Eddy]

The Eone has 'bridged' XLR outputs...

Thank you.

se

 

[dprimary]

I did a quick test using AKG headphones and the standard three wire AKG K240 cable that I have had my techs make into 5' lengths which allow me to open the mini XLR shell and measure the crosstalk. In my quick test that was limited by the range of my RMS meter the separation was over 50dB using both 1k sine waves and 20-20k pink noise. I think you must have something else wrong, Every time I have heard large amount of crosstalk on headphones or IEMs either something was mis-wired the plug are jack was damaged or not seating fully. 

 

[lister]

  I did a quick test using AKG headphones and the standard three wire AKG K240 cable that I have had my techs make into 5' lengths which allow me to open the mini XLR shell and measure the crosstalk. In my quick test that was limited by the range of my RMS meter the separation was over 50dB using both 1k sine waves and 20-20k pink noise. I think you must have something else wrong, Every time I have heard large amount of crosstalk on headphones or IEMs either something was mis-wired the plug are jack was damaged or not seating fully. 

 
Thanks for sharing.  If the experiment is done right, then it appears that the AKG K240 cabling is in good shape.
Is it possible for you to share a picture of your experiment?  Or a schematic sketch?
 
Also, please note my responses above to Steve Eddy.  I was not questioning XLR cables at all  Those cables usually have separate ground return wires for each L&R.  It's not worth your time to find crosstalk in those type of cables, because it is unlikely.
Do you have any earphones with normal cables (as in not XLR)?
 
If you see the two pictures I've posted, I was more concern with earbuds/IEMs.  And you don't need to re-work the wiring to run my experiments.  All you need is let your ears be the judge.

 

[lister]

Dear all,
 
I did find an article regarding crosstalk in headphones (from SOS)...
http://www.soundonsound.com/sos/jul13/articles/qanda-0713-1.htm
 
A quote from this article... "SOS Technical Editor Hugh Robjohns replies: Yes, this crosstalk issue is well known..."
 
Isn't it absurd that we build amplifiers with output channel separation > 90dB, and yet it all get ruined by our headphone cables (~30 to 50dB)??