[Steve Eddy]

A few months or so ago I thought I'd evaluate crosstalk as it related to the common contact resistance of the TRS jack/plug combination and found that it results in a significant amount of crosstalk.

Today I thought I'd evaluate crosstalk is it relates to capacitive/inductive coupling in the headphone cable.

I started by pulling 24 feet of telephone station wire off a spool I had on hand. This wire has four strands of PVC insulated 24 gauge solid core wire in a hard PVC jacket.

One pair of wires was used for the driven half of the cable. One end was connected to the speaker terminals on a JVC FS-2000 mini system and the other end to an 8 ohm resistive load.

The other pair had a 150 ohm resistor tied across one end to serve as an amplifier's output impedance. While most amps won't have such a high output impedance, save perhaps some tube amps, I was wanting to go for a worst case scenario.

The other end was tied to a 1/8" TRS jack with one wire being attached to both tip and ring contacts and the other to the sleeve contact.

I plugged in some cheapie Radio Shack Nova 43 headphones, fired up the FS-2000 and started turning up the volume.

Unlike the previous test, I wasn't treated to Wynona's Big Brown Beaver or anything else for that matter. Even with the volume turned up all the way and the FS-2000 slightly clipping, there was nothing to be heard.

So it doesn't appear that crosstalk within a four conductor headphone cable is anything to worry about.

se

 

[scootermafia]

Let me get this straight. You have a 4 wire cable. You have one of the pairs from a speaker terminal pair going to an 8 ohm load. The other pair, which is next to the speaker-8 ohm load pair in the cable, is running from a 150 ohm resistor, to a headphone jack (running left/right channels merged). I don't see exactly how you would hear anything, it's not like a significant current could be going from the 150ohm resistor-headphone. My concern is that when both pairs have currents going through them, common mode noise rejection (twisting the pairs together) might not be enough to separate the two pairs. The currents in each pair going back and forth - the left and right pairs may still affect each other when signal is going through both. And just because you can't hear music in those headphones, doesn't guarantee that this effect in an actual headphone cable won't do something to the detail of the music. I'm afraid I'm not 100% sold on this one.

 

[Steve Eddy]

Quote:

Originally Posted by scootermafia /img/forum/go_quote.gif Let me get this straight. You have a 4 wire cable.

Yes. With the wires in intimate contact with each other.

Quote:

You have one of the pairs from a speaker terminal pair going to an 8 ohm load.

Yes. Specifically a resistor load, not a loudspeaker (need to maintain silence in the room).

Quote:

The other pair, which is next to the speaker-8 ohm load pair in the cable, is running from a 150 ohm resistor, to a headphone jack (running left/right channels merged).

Yes.

Quote:

I don't see exactly how you would hear anything, it's not like a significant current could be going from the 150ohm resistor-headphone.

Exactly.

Because the only means by which one channel can couple to the other in the cable is by way of capacitive and/or inductive coupling.

Quote:

My concern is that when both pairs have currents going through them, common mode noise rejection (twisting the pairs together) might not be enough to separate the two pairs. The currents in each pair going back and forth - the left and right pairs may still affect each other when signal is going through both.

There's no need to have currents flowing through both pairs. If the driven pair doesn't couple to the un-driven pair sufficiently to produce anything audible, then nothing will couple sufficiently with both pairs driven.

Quote:

And just because you can't hear music in those headphones, doesn't guarantee that this effect in an actual headphone cable won't do something to the detail of the music.

The only thing that could possibly effect any detail of the music would be if there was audible coupling between channels. If there's no audible coupling between channels, then there's nothing that can possibly effect detail.

se

 

[Punnisher]

You should try the same test with a four conductor, enamel coated copper cable.

The total thickness of the enamel between two conductors is probably much less than a thousandth of an inch. I have always detested this kind of cable (especially when it comes to maintenance) and I wonder if the extremely close proximity of these conductors would contribute to crosstalk.

 

[Steve Eddy]

Quote:

Originally Posted by Punnisher /img/forum/go_quote.gif You should try the same test with a four conductor, enamel coated copper cable. The total thickness of the enamel between two conductors is probably much less than a thousandth of an inch. I have always detested this kind of cable (especially when it comes to maintenance) and I wonder if the extremely close proximity of these conductors would contribute to crosstalk.

Ok.

I'll try and swing by Radio Shack tomorrow and pick up some magnet wire.

Did you have any particular gauge wire in mind?

se

 

[Punnisher]

Not in particular. Maybe the most common headphone gauge which could be 28awg or even smaller.

Not that it matters much.

The type of cable I was referring to is the most common headphone cable type. Usually three stranded conductors, woven with fiberglass, each coated in enamel, twisted together and finally coated with a jacket.

Though magnet wire should replicate that fairly well. Plus the fact that it's solid means you don't need it to be as thick. Just be sure to twist it tightly enough but not so tight as to damage the enamel and create more crosstalk.

 

[Steve Eddy]

Okie doke.

I don't think I'll twist it though. Instead, I think I'll get a 100 foot spool, set up a couple of sawhorses with some dowels on them 24 feet apart, and do four runs of it then bundle the wires together with Teflon plumber's tape.

se

 

[Steve Eddy]

Ok, it's going to be 30 gauge.

Decided to grab a pizza for dinner and the pizza joint's near Radio Shack so I stopped in and grabbed the wire. They only sell magnet wire in a three pack that has 40 feet of 22 gauge, 75 feet of 26 gauge and 100 feet of 30 gauge.

se

 

[Anonanimal]

/subscribed. Very interesting tests and results! Thanks for doing this and I'm looking forward to what you find out.

 

[Steve Eddy]

Well, I really wasn't too keen on the notion of taping up the wires but fortunately I happened to have some 3/64" heatshrink that's small enough that it shrinks up nice and snug on four 30 gauge strands.

Got out the sawhorses and dowels and ran four 24 foot strands of 30 gauge wire and inserted them into the heatshrink using 1 foot lengths at a time then shrunk it all up.

Have some running around to do but should be able to run the test later this afternoon.

se

 

[scootermafia]

I just don't see how just because you can't hear an audio signal, that the inductive effects between wires couldn't affect each other. Like, it may not be immediately audible, but it might be able to change the sound signal.

 

[Steve Eddy]

Quote:

Originally Posted by scootermafia /img/forum/go_quote.gif I just don't see how just because you can't hear an audio signal, that the inductive effects between wires couldn't affect each other. Like, it may not be immediately audible, but it might be able to change the sound signal.

The only effect would be one channel's signal bleeding into the other channel and vice versa by way of inductive and/or capacitive coupling. If that crosstalk isn't of sufficient magnitude to be audible, then by definition it can have no effect on the sound of anything.

se

 

[IPodPJ]

And what driver/transducer were you playing the audio through? What was the sensitivity rating? What was the wattage?

Edit: Oh, you already mentioned the headphones. "I plugged in some cheapie Radio Shack Nova 43 headphones". I'd like to know what the sensitivity rating is, it's impedance, how much power it can handle, and how much you were feeding it.

 

[scootermafia]

I was going to say, with a low-end DAC/amp/headphones you cannot hear subtleties in music anyway, they get squashed out. A good example is at the intro of the song Back in Black by AC/DC, in the first few seconds the lead singer says "awwwwww" like "awwwww yeahhh" or however you want to describe it. You can't hear that with a cheap DAC, I found. It's just a minute detail that is not very clear to hear in many cases. If anyone should be doing this test, it's me.

Right, and in response to what PJ said, this effect may not be audible in all setups/configurations.

 

[Steve Eddy]

Quote:

Originally Posted by IPodPJ /img/forum/go_quote.gif Edit: Oh, you already mentioned the headphones. "I plugged in some cheapie Radio Shack Nova 43 headphones". I'd like to know what the sensitivity rating is...

No idea what their sensitivity rating is. They play quite loud from the output of my iPod Nano though.

Quote:

...it's impedance...

32 ohms.

Quote:

...how much power it can handle...

What's that have to do with anything relevant here?

Quote:

...and how much you were feeding it.

I was feeding it as much as any crosstalk in the cable was capable of feeding it.

se