How Connectors, Cables, or Solders sound?

[Earfonia]

Another 2 ohms cable:
https://www.linum.dk/products/linum-g2-music/

Multi-driver IEMs are usually those that affected by Cable and headphone output Impedance. Usually less so on dynamic driver earphones. But usually at 2 ohms the affect to tonality is minimal (probably might be audible in some cases).

 

[Julf]

I found the article on Innerfidelity that discussed the effects of cable resistance

Yes, high resistance does have an effect, even more so on speaker cables, especially if the headphones / speakers have a very uneven impedance curve. There is a simple solution - cables of sufficient thickness / gauge. No magic.

 

[Fluffy]

Of course no magic. It's pure (measureable and explainable) physics.

 

[solderdude]

Impedance mismatch between amp and headphones

There is nothing to match between output resistance, cable impedance and headphone impedance.


re cable resistance:

Let me give a real world example of the famed AQ Nighthawk (measurements of the cable here).
It comes with 2 cables which sound different yet measure exactly the same in FR (of course)
The 'trick' they applied (that's the way I see it) to promote their expensive cables is to have 2 of them that differ enough to make people 'believe' the difference in materials matters.

AQ NH = 25 Ohm.
return wire (silver) = 0.5 Ohm
return wire (gold) = 1.6 Ohm

the signal wires are also the same resistance. This results in a level difference of 0.8dB between the cables.

BUT now we are going to take crosstalk into account.
For the silver cable its -34dB
For the (inferior acc. to AQ) gold cable = -24dB !

When one could use a 4 wire cable connected in the 6.3mm plug) that crosstalk could be way down -50dB
With a 'balanced' cable well below say -100dB.

 

[Fluffy]

There is nothing to match between output resistance, cable impedance and headphone impedance.

Matching impedances can significantly impact bass response on most dynamic headphones. My Focal Clear has a very strong primary driver resonance which make the impedance curve look like this (from Innerfidelity):

Connecting them to an amplifier with high output impedance will result in elevated levels of bass response, as shown by innerfidelity (Simaudio Moon NEO 430HA in purple and Bottlehead Crack with about 120 Ohm output impedance in orange).

:

And also DIY-Audio-Heaven (red is 0.2Ω, green 120Ω) :

And RAA:

 

[solderdude]

That's correct but it does not have anything to do with matching.
When matching the source and load impedance are the same and the cable impedance too.
Here you are just seeing the effect of voltage division due to a higher source resistance.

 

[Julf]

Matching impedances can significantly impact bass response on most dynamic headphones.

I don't think of ensuring the amp has low enough output impedance. To me "matching impedance" implies keeping the impedance the same at input, output and cable to avoid reflections in the cable (think of 50 and 75 Ω RF connections and 600 Ω phone lines).

 

[Fluffy]

Well maybe I was using the wrong word (English is not my native tongue). My point is that output impedance is an important factor to consider, and could have an impact on frequency response.

I don't really know how reflections affect the sound. Can anybody explain?

 

[Julf]

Well maybe I was using the wrong word (English is not my native tongue). My point is that output impedance is an important factor to consider, and could have an impact on frequency response.

Yes, that is correct, if just isn't "impedance matching".

I don't really know how reflections affect the sound. Can anybody explain?

They don't, not at audio frequencies and cable lengths less than a football field.

 

[Fluffy]

I believe you, but still I would like to understand what is the physical effect, never mind in the audio band.

And how would you call it instead of impedance matching?

 

[Julf]

I believe you, but still I would like to understand what is the physical effect, never mind in the audio band.

Reflections of signals on conducting lines

And how would you call it instead of impedance matching?

Simply "low enough output impedance", as it doesn't need to be matched, just low enough.

 

[solderdude]

Cable impedance is negligible compared to the effects seen with some headphones when driven through an unusally high output resistance.
That's what the discussion originated from.

By the way... the RAA plot is a bit suspect. All the other measurements have the impedance peak around 40Hz (Tyll's earlier version 50Hz) but the RAA measurement is around 110Hz.

 

[Cbdb2]

The high output impedance amp thing is just confusing this thread, about cables. Why would anyone make a high impedance headphone amp?

 

[Earfonia]

The high output impedance amp thing is just confusing this thread, about cables. Why would anyone make a high impedance headphone amp?

This article probably useful:
https://www.innerfidelity.com/content/musings-headphone-amplifier-output-impedance

Basically for some headphones higher output impedance can be used to optimize damping.

 

[Julf]

Basically for some headphones higher output impedance can be used to optimize damping.

For "optimize" read "reduce", the result will be larger frequency response variations caused by the uneven impedance curve of the headphones.

 

[Earfonia]

For "optimize" read "reduce", the result will be larger frequency response variations caused by the uneven impedance curve of the headphones.

Yes, that's true when the impedance curve of the headphone is highly uneven. I did test Audio-Technica headphone amplifier, Audio-Technica AT-HA5050H, that has unique multiple headphone output with variety of output impedance: 0.1 ohm, 33 ohm, 82 ohm, and 120 ohm. I tried with Beyerdynamic T1 and Sennheiser HD800. From what I observed, the frequency response variations are not great. I would say audible but only mild variations from the lowest to the highest output impedance. And for HD800 I do prefer to use the higher output impedance, the 33 ohm, as it seems to tame the treble peak a little bit. Transient was also affected, the lower output impedance sounds best for maximum transient. That's what I observed from listening.

The TEAC HA-501 also has output impedance selector for the headphone output:
http://audio.teac.com/product/ha-501/

Some engineers see it as a feature to tweak the sound quality.

From my experience so far, single dynamic driver headphones and IEMs are less susceptible to amplifier high output impedance. But for multi drivers IEMs the effect can be pretty bad, and they often a lot more susceptible to high output impedance.

 

[Julf]

From my experience so far, single dynamic driver headphones and IEMs are less susceptible to amplifier high output impedance. But for multi drivers IEMs the effect can be pretty bad, and they often a lot more susceptible to high output impedance.

Indeed. But are there any cases where high output impedance would actually be beneficial?

 

[solderdude]

And for HD800 I do prefer to use the higher output impedance, the 33 ohm, as it seems to tame the treble peak a little bit. Transient was also affected, the lower output impedance sounds best for maximum transient. That's what I observed from listening.

HD800S on 0.2 Ohm and 120 Ohm:

Only a 1dB lower frequency boost. Treble peak is not affected at all. I also made transient plots (square-wave and CSD) which showed no differences other than the very mild lows boost. These plots were not posted.

b.t.w. I have a short silver cable and the original cable (both different connectors as well of course) and there are absolutely no measurable nor audible differences.

Taking a headphone of the head (or test rig) and placing it back usually results in substantially measurable differences that are considered audible.
This is at least one of the reasons why cable swapping and listening again (often the same song) is a terrible flawed way to go about this but unfortunately can not be avoided.

It is possible you observed and heard this though. Let's just say it is not reflected in the measurements.

 

[Earfonia]

Indeed. But are there any cases where high output impedance would actually be beneficial?

I guess we have to ask the engineers who designed the AT HA-5050H and TEAC HA-501 for that.

It is possible you observed and heard this though. Let's just say it is not reflected in the measurements.

Agree. Could possibly just observational bias. As I said the effect was very mild. +1 dB on the bass could also be translated by our hearing as -1 dB on the treble as our hearing hear the whole tonal balance. And 1 dB difference is audible when we listen carefully in a quiet room. Or probably the slight reduction of the overall listening level. Difficult to accurately manual volume matching from low to high output impedance. Anyway, 'I think' I heard very small differences when comparing the outputs with different output impedance. Nothing significant. How to translate it in measurement, or was it totally perception bias, I have no idea. I was not trying to convince myself which one better when I was comparing the outputs, so from that point of view I didn't have perception bias. Just sharing my experience.