Custom Cable Question about AWG

I am looking to build my own balanced cable for Senn HD6xxxxxx, but everywhere I read seems to be inconsistent with what the AWG of the wire should be.

I read 26 AWG is ideal
But then again so is 21 AWG
But most cables are 16 AWG?

I’m confused

One good manufacturer uses Cardas 4 X 24 AWG or Mogami 2893 4x26 AWG.
Probably they are very good cables and don’t need to worry about anything else.
Expect making all the soldering and connector contacts just right…


gauge determines how much juice can go through it. there are online gauge calculators to determine over all gauge if you combine small gauges in a braid. heavier gauge weighs more, i wouldn’t “overkill” hp’s or iem’s unless your good with the added weight.


To minimize transmission losses, you want as much conductor cross-section as reasonable (= lower AWG/higher mm²). There is also the “skin effect”, so for Audio ideally the individual wires are below 19 AWG (~ 0.65 mm²). At that cross section, you can push almost 2 Ampere (which would grill your headphones after blowing your eardrums out).

The more wires the conductor is made of, the more flexible the cable is.
I have some silicone mantled cable that is 24x 0.02mm², super flexible stuff!

Handy chart:


on a further note, to determine how much gauge you need. take max rated power of iem or hp, then look up a current value by gauge chart. there are many online. most iem’s do not need more then 26 gauge combined and most headphones do not need more then 22 gauge.

I’ve been doing research on building cables, and star quad Mogami cable is often suggested and that’s 26 AWG. You can also consider how the original stock HD600 cable was 26 or 28 AWG.

But personally, I’d go with 24 AWG or larger since quality cable at that gauge is readily available. If you’re doing a fancy braided cable, you may as well go 20 or 22

Anyway I recently bought a 10 ft 25AWG XLR extension. And I was curious if I would see a measurable difference on my EARS rig if I used power hungry headphone. It turns I could. Adding the extension reduced the measured FR response at low frequencies by about 0.07 dB.

LCD-2.2 PF, set to 96 dB at 300 Hz.