DIY headphone cable: which is the best rule ?

[ygg-it]

I'm going to recable my akg271.

So which is the best doing it?

Here some my thoughts, but please tell me if i'm right and add any new one!!!


1) cable lenght must be as shorter as possible

2) cable must not have one common shield, but two separate ground wires

3) twisted cables is better than shielded (?)

4) multi wires twisted cable is better than two wires twisted cable

5) multi diameter twisted wires is better than same diameter twisted wires

6) big diameter wires is better


So I'm thinking to build:

1) 2 meter max lengh cable
2) phisically separate wiring for each channel
3) for each channel
right (+): twisting 1 mmsq and 0.2 mmsq
right (-): twisting 1 mmsq and 0.2 mmsq
right (+) twisted with right (-)
left (+): twisting 1 mmsq and 0.2 mmsq
left (-): twisting 1 mmsq and 0.2 mmsq
left (+) twisted with left (-)
right and left NOT TWISTED each other but run in parallel

What you think?

 

[FallenAngel]

1) Meh, never noticed a difference, pure preference; just don't use extension adapters (joins the grounds of unbalanced cables).

2) Yep

3) Would be nice to verify, I haven't noticed a difference in using a shielded vs unshielded cable for headphones.

4) Preference again; I prefer solid-core for everything, 26awg is great for headphones.

5) Getting into "audiophile marketing" bulls--t here, sorry. Having the wires wires of different size means the signal will be able to go through one faster than the other, might not be something you want.

6) Nope

Build:
1) Make it the length you will use
2) Would be nice, yes
3) Sounds good, but I personally prefer not to have mix wire sizes for same signal.

 

[Fungi]

I do know some people swear by Litz braiding. I can't say through experience or measurement but it's supposed to kill noise and crosstalk or something.

 

[qusp]

yes litz braiding the grounds with the signals helps to kill RF and that is why there is not so much need for a shield as a shield surrounds a cable connected to ground and a litz braid effectively surrounds the signal with ground too. but over long lengths can cause increased capacitance. as long as you cable isn't too long it would be fine though. I personally would braid left and right with both grounds together until the Y split and then cover each twisted pair with shrink up to the drivers

this is just my preference, there are many different methods for making a HP cable, but as you are making a much longer cable than say; an LOD or mini to mini. there are a few more forces that come into play. with braiding its capacitance, with parallel it becomes parasitic capacitanceThere are several methods of making a headphone cable but this one seems to be the most popular. and for good reason; it sounds great IMO

with HP cables thicker is absolutely NOT better.
as fallen angel said, make it whatever length is best for your setup, you dont want to be tethered to the amp without having freedom to move around comfortably. and yes the multi diameter wire thing is all hype.

ummm yes seperate wiring for each signal is recommended

 

[qusp]

double post sorry

 

[JamesL]

I would just use two thin twisted pairs(24-28awg), teflon dielectric, and perhaps a cotton or teflon tubing over the twisted pair for mechanical support.

As the headphone cable is carrying a low impedance signal, albeit not the same as speaker cables, it's not as sensitive to RFI and capacitance as interconnects are.
A litz braid or twisted ground wire is mediocre at shielding, when compared with a proper braided, stranded, or foil shield that is not conducting any current.
A thinner dielectric or a tighter braid will increase capacitance, but lower inductance, and increase noise rejection. A thicker dielectric, thinner wire, or looser braid will lower capacitance, at the cost of the other prior mentioned.
Too thin of a wire, and you'll be imposing a somewhat significant resistance for a 3m cable(about 1ohm for 31awg). Because impedance varies at varying frequencies, dc resistance may change the fr of the headphones.

Note, that any scientific differences derived from different cables are extremely small, and arguably irrelevant.

If it was me, I'd be more concerned about getting the better built plug, the nicer heatshrink, right sized sleeving, etc.

 

[rds]

Keep in mind that litz is derived from a German word and it effectively means 'bundled' or 'braided'. So the term 'litz braid' is redundant and doesn't really specify anything. You might as well just say braided wire.
On the other hand litz wire is a real specialty type wire that is used to reduce the skin effect for transmitting high frequency signals (way higher than audio).
Twisted pairs do reject EMI by influencing the ground equally and opposite to the hot wire.
So these concepts get confused a lot.
Shielded twisted pairs is theoretically the best way to go. Whether it's significant in transmitting audio signals is another question.

 

[NelsonVandal]

Quote:

Originally Posted by ygg-it /img/forum/go_quote.gif 1) cable lenght must be as shorter as possible 2) cable must not have one common shield, but two separate ground wires 3) twisted cables is better than shielded (?) 4) multi wires twisted cable is better than two wires twisted cable 5) multi diameter twisted wires is better than same diameter twisted wires 6) big diameter wires is better What you think?

1) agree, low impedance is essential
2) why use a shield?
3, 4, 5) I'm not sure, I'll pass on those questions
6) agree, low impedance is essential

I'm using a short twisted 0.75 mm2 silicon rubber cable. This kind of cable is used in RC cars. Thick as it is, it's yet very compliant. It sounds better than the original HD650 cable and better than 24 AWG SPC.

Why isn't impedance important in headphone cables? What's the theory of thinner being better? As far as I know the only proof of difference in blind tests has to to with impedance and not chrystaline structure, purity, special materials etc.

I've found while fiddling with amplifiers that low impedance for the headphones return currents is of great importance, and the cable is part of it. Higher ground channel impedance gives a wider and leaner sound. Impressive at first but tiresome in the long run. I'm no electronic wiz, and there's a great risk I'm wrong. These are just my findings.

If you think the cable alters the sound, there's no logic in saying the cable length is unimportant? It's very contradictive.

 

[bhjazz]

These are all good quetions, but don't obsess with them too much. Even if you try two cables that are built the same way but by different manufacturers, they'll sound different. Instead, head out to PartsConnexion or your wire dealer of choice and get some wire, build a few, and see which one you like best.