I need to solder a new cable on headphones for someone, and would like to keep a spare cable ready for my AKG 242's. So em, what cables work best for headphones?
When I take apart any kind of headphones, they usually use very thin wires without insulation but with a colored enamel. How is this kind named?
When I take apart any kind of headphones, they usually use very thin wires without insulation but with a colored enamel. How is this kind named?
Uh... enameled wire?
Headphone cables are a bit of a science unto themselves. They need to be light, flexible, with low microphonics, yet robust. Your best bet may be some kind of microphone cabling with foamed PU sleeve in starquad configuration.
Note that the ground returns of left and right driver are to be kept separate up to the plug.
Headphone cables are a bit of a science unto themselves. They need to be light, flexible, with low microphonics, yet robust. Your best bet may be some kind of microphone cabling with foamed PU sleeve in starquad configuration.
Note that the ground returns of left and right driver are to be kept separate up to the plug.
Just a thought, POCC cables are good for wiring from the amp out to the binding post. Why not to give it a try for the headphone ?
My thought is that headphones cable is (or must be) a part from the overall system.
I can share only my small experience:
I decided to replace the stock cable of Beyerdynamic DT880 250 Ohm - not for fun but because I'd like to have headphones with balanced input. I had at hand some good silver plated cable with teflon (PTFE) insulation, made by Habia (habia.com).
However, even though better cable's quality, the result was disastrous - big peaking in the mids and overall unnatural sound. Stock cable was returned again.
On my second headphones - Beyerdynamic DT880 600 Ohm I did the same operation but this time with more ordinary copper cable again with PTFE insulation (some Russian production I think). This time result was very good - the tonal balance was retained, just the headphones' input became balanced.
So, the moral (for me) is that even with very high quality cables the results are unpredictable.
I can share only my small experience:
I decided to replace the stock cable of Beyerdynamic DT880 250 Ohm - not for fun but because I'd like to have headphones with balanced input. I had at hand some good silver plated cable with teflon (PTFE) insulation, made by Habia (habia.com).
However, even though better cable's quality, the result was disastrous - big peaking in the mids and overall unnatural sound. Stock cable was returned again.
On my second headphones - Beyerdynamic DT880 600 Ohm I did the same operation but this time with more ordinary copper cable again with PTFE insulation (some Russian production I think). This time result was very good - the tonal balance was retained, just the headphones' input became balanced.
So, the moral (for me) is that even with very high quality cables the results are unpredictable.
okay, interesting. where do you buy such kind of specialised cabling?
my local reseller says normal copper cabling is equally good, but is thicker because it doesn't rely on enameling for electric insulation
my local reseller says normal copper cabling is equally good, but is thicker because it doesn't rely on enameling for electric insulation
to avoid ground loops?
While not exactly the same as what you're describing, litz wire, as used by ham radio enthusiasts, should fit your needs nicely.
If you search something like "litz 100/46" on eBay you should get a few hits. Those dimensions will give you a wire with roughly the same conductance as 24AWG regular stranded wire, but be very thin and supple. You can of course go for an even thinner wire if you want to, like 60/46 (60 strands of individually insulated 46AWG wire)
If you search something like "litz 100/46" on eBay you should get a few hits. Those dimensions will give you a wire with roughly the same conductance as 24AWG regular stranded wire, but be very thin and supple. You can of course go for an even thinner wire if you want to, like 60/46 (60 strands of individually insulated 46AWG wire)
OK, thats a good tip. "litz" gives a bunch of results on Ebay.
earlier I was digging. with the keywoard "PTFE" or "teflon" I mostly find silver-plated copper cabling, rather expensive.
One page describes the litz cable as following, with its advantages and disadvantages:
I assume its possible but opinions will probably vary a lot 😀
earlier I was digging. with the keywoard "PTFE" or "teflon" I mostly find silver-plated copper cabling, rather expensive.
One page describes the litz cable as following, with its advantages and disadvantages:
Can twisted pairs can be used for headphone cabling?
I assume its possible but opinions will probably vary a lot 😀
Unless you want to go into the whole mumbo-jumbo of cables having a 'sound', pretty much anything goes, as long as RLC values are kept within reason.
It's more about functionality and aesthetics, really. If soft and supple is what you want, litz is a good option. If strong and durable is what you want, then something more conventional, like Mogami W2893, would do the trick.
I added a couple of examples. The first one is litz with a cotton sheath arranged in a square braid to form an interconnect. The second one is Mogami quad with the outer rubber sheath removed and replaced (up to the Y-split) with a woven polyester mantle.
Attachments
I can't understand the ignorance of the mechanical aspect - most 3rd party or handmade cable are stiffer, heavier and have rougher finish than most manu supplied headphone cable - all add to microphonics - a real sound distraction that no LCR improvement is likely to make up for
Agreed, headphone cables are about 80% about mechanics. (Hence why I suggested microphone cabling, where you can get fairly lightweight yet robust and non-microphonic material.) Electrically you tend to be fine if resistance does not add up to more than 1-5% (or maybe 10%) of nominal driver impedance, and both drivers' ground returns are kept separate right up to the plug.
The latter is because a shared return, in measurement terms, degrades crosstalk. To be more precise, it attenuates the L+R (in-phase) components as opposed to L-R (out of phase) - essentially a classic stereo widening effect. It's not a big deal at mid-high driver impedances like 100 ohms up since headphone cables rarely tend to exceed about 1 ohm and nobody really cares about -40 dB worth of crosstalk either in or out of phase - but go down to 16 ohms, and it's a different story entirely.
Assume we have two 16 ohm drivers, and cabling with 1 ohm per run.
Case 1: Separate ground return.
* (L+R) sees (1 + 16 + 1) / 2 ohms = 9 ohms
* (L-R) sees (1 + 16 + 1) * 2 ohms = 36 ohms
Case 2: Shared ground return - let's combine our returns directly behind the drivers while retaining the same cable (unrealistic but good for illustration).
* (L+R) sees ((1 + 16) / 2) + 0.5 ohms = 9 ohms
* (L-R) sees (1 + 16) * 2 ohms = 34 ohms
Note how in case 1, the same impedances are involved, just one time in parallel and one time in series, hence the difference of exactly a factor 4.
In case 2, nothing much has changed for L+R. L-R, however, has just become louder. Instead of 32 / 36 of total power, the drivers receive 32 / 34, upping power by 36/34 = 5.9% or +0.5 dB.
In practice, a 3-conductor cable would be used instead, which would double shared return resistance and make the disparity correspondingly worse. Now can you hear a ~10% stereo expansion? Absolutely!
If you do the math, it is also possible to determine an estimate of shared return resistance by measuring resistance of (L to return), (R to return) and (L to R) at the plug. A good, well-calibrated multimeter is required, however.
Since I did this, my Rockbox sound setting preset for Soundmagic E10s includes a "custom" channel configuration with stereo width at 90%...
Note that this simple kind of compensation only works for loads with nearly-ohmic impedance responses, as dynamic in-ears thankfully tend to have. In general, the crosstalk level would be frequency-dependent.
Interestingly enough, it is not only higher driver impedance that makes this issue less critical - additional output series impedance has the same effect.
The latter is because a shared return, in measurement terms, degrades crosstalk. To be more precise, it attenuates the L+R (in-phase) components as opposed to L-R (out of phase) - essentially a classic stereo widening effect. It's not a big deal at mid-high driver impedances like 100 ohms up since headphone cables rarely tend to exceed about 1 ohm and nobody really cares about -40 dB worth of crosstalk either in or out of phase - but go down to 16 ohms, and it's a different story entirely.
Assume we have two 16 ohm drivers, and cabling with 1 ohm per run.
Case 1: Separate ground return.
* (L+R) sees (1 + 16 + 1) / 2 ohms = 9 ohms
* (L-R) sees (1 + 16 + 1) * 2 ohms = 36 ohms
Case 2: Shared ground return - let's combine our returns directly behind the drivers while retaining the same cable (unrealistic but good for illustration).
* (L+R) sees ((1 + 16) / 2) + 0.5 ohms = 9 ohms
* (L-R) sees (1 + 16) * 2 ohms = 34 ohms
Note how in case 1, the same impedances are involved, just one time in parallel and one time in series, hence the difference of exactly a factor 4.
In case 2, nothing much has changed for L+R. L-R, however, has just become louder. Instead of 32 / 36 of total power, the drivers receive 32 / 34, upping power by 36/34 = 5.9% or +0.5 dB.
In practice, a 3-conductor cable would be used instead, which would double shared return resistance and make the disparity correspondingly worse. Now can you hear a ~10% stereo expansion? Absolutely!
If you do the math, it is also possible to determine an estimate of shared return resistance by measuring resistance of (L to return), (R to return) and (L to R) at the plug. A good, well-calibrated multimeter is required, however.
Since I did this, my Rockbox sound setting preset for Soundmagic E10s includes a "custom" channel configuration with stereo width at 90%...
Note that this simple kind of compensation only works for loads with nearly-ohmic impedance responses, as dynamic in-ears thankfully tend to have. In general, the crosstalk level would be frequency-dependent.
Interestingly enough, it is not only higher driver impedance that makes this issue less critical - additional output series impedance has the same effect.
Last edited:
- Status
- Not open for further replies.