Interconnects - shielded or unshielded?

[YNWOAN]

So, here is a potentially contentious thread, not least because it is about interconnect cables and the subjective experience of them .

I should add that, whilst I quite appreciate that some/many feel ‘wire is wire’, the premise of my following question is not ‘can interconnects make a difference?’ But, even more esoterically, what sonic difference is there between shielded and unshielded interconnects - and, therefore, by inference, what is the sonic impact of interconnect shielding?

The following excludes the use of balanced cabling as it lies outside of my experience.

Over the last few months I have revisited interconnect performance (sound? - or rather, their impact upon the subjective sound) and experimented further (than my previous experiments). One of the most common design features, and one of the less ‘hypothetical’ advantages, that many interconnect cable manufacturers (can there really be such a thing as an interconnect/cable designer?) is shielding. Sometimes this shielding is a semi-conductive plastic film, sometimes, aluminised foil, sometimes a metal braid and, quite often, a mixture of two of these. Occasionally, multiple braided shields will even be used. In the majority of cases, at least when the shield does not form the signal return, the shield is connected at one end (only) to signal ground (conceptually the end of lowest impedance). Completely unshielded interconnects do exist but are rarer than one may think.

So, my question is this; do fellow Fishers with an experience of different interconnects feel that shielding of interconnects carries a sonic price/disadvantage?

 

[Tony L]

I defy anyone with a genuinely good audio system not to be able to hear the difference between say Kimber PBJ (un-screened PTFE coated open weave) and a decent screened coax. They do sound different, though as with any difference it is subjective and folk will just pick what they like.

There is also the idea of just screening at one end, e.g. van den Hul D102/III, which is a twin-core cable with an outer screen that is only connected at the source end.

PS There is a very noticeable sonic difference between say PBJ and D102, but I’d not like to say it was just down to screening as there are many other differences, plus they’ll measure differently, and I’m personally more inclined to believe sonic character relates to measurable electrical parameters.

 

[YNWOAN]

Well yes, I quite agree - with all of that.

Playing devil’s advocate for a moment, I believe Kimber claim that the woven quality of their cables imparts an element of shielding (though this is far from being an uncontentious claim).

 

[Tony L]

No idea about that, though I’ve never heard PBJ hum (I have a 1m pair in my cable tub just in case some objectivist evangelical type calls around as you really need to be deaf as a post not to hear the character of that cable!).

 

[JemHayward]

I think it may well depend on what is going down the wire, as I can imagine that the small signal from a low output MC cartridge is more likely to be affected than the signal between preamp and power amp. Also, screened cables tend to have more capacitance, so it would be difficult to compare a screened cable with an unscreened cable, if you think that capacitance has any effect. In recent years I tried comparing two balanced cables (between DAC and power amp) of different lengths and construction, and found absolutely no detectable difference at all, so I think that going fully balanced is a way to get rid of the problem once and for all.
When I had my hugely complex active Isobarik setup I had unscreened cables to the power amps (5A mains lead) but found I had to use screened mains cables, which kept some odd interference from getting into my Naim preamp.

 

[YNWOAN]

All my trials have been between my phonostage and my pre-amp.

Van Damme do a cable that is available as a low, mid and high (relatively) capacitance version. During earlier trials I made up identical cables using the high and low capacitance versions of the cable (coax construction). I used the same connectors, solder etc. Differences between these cables were absolutely tiny/non existent which rather suggests that, if only using a metre, capacitance has little, or no, impact.

 

[Coda II]

Sorry to be simplistic, but doesn't screening suggest that what is inside is being screened from something outside (and vice versa in eg. the case of satellite dish coax)? So in an environment where there is no external influence to be screened from, then screening would confer no advantage and it would be possible to try and ascertain whether two otherwise identical cables (one with screening one without) behaved any differently, whether the screening was in any way detrimental.
As an aside, in an ongoing effort to reduce hum from a pair of mono-blocks switching to screened power cables has been beneficial (when DC blockers were not) whether this is purely down to the screening I don't know.

 

[JemHayward]

I think that unless either the phonostage or the preamp have very odd impedance characteristics cable capacitance, for a short cable, will be irrelevant. Even before Naim started selling boutique cables, they did suggest keeping the low level signal cables short, and high level speaker cables longer as the best compromise, but they did make me up a 4m Snaic to use between my NAC82 and my NAXO3 and I couldn't hear any difference between that and the standard cable (and it was charged at the same price too!).

 

[YNWOAN]

@ Coda II, If the hum from your amps was mechanical, the transformers themselves humming (rather than a hum coming from the speakers) I doubt a screened mains cable would do anything at all to reduce the hum (I don’t see how it could in fact). In a similar attempt to reduce transformer hum I also tried a D.C. blocker and found it made no difference. However, I later tried a much larger D.C. blocker and that did, entirely, stop the hum.

 

[Coda II]

If the hum from your amps was mechanical, the transformers themselves humming (rather than a hum coming from the speakers) I doubt a screened mains cable would do anything at all to reduce the hum (I don’t see how it could in fact). In a similar attempt to reduce transformer hum I also tried a D.C. blocker and found it made no difference. However, I later tried a much larger D.C. blocker and that did, entirely, stop the hum.

Yes transformer hum and equally I'm puzzled as to why screening would affect it. What was the larger blocker that worked?

 

[cj66]

When messing about with DIY cables I've always had a preference for pseudo balanced as per the VHD102iii arrangement Tony described above.

 

[John Phillips]

Addressing just basic engineering, to take an analogue signal from one box to another needs a signal and a return connection. If you want to keep the high dynamic range possible with modern audio I would always screen them. Electromagnetic interference is a bit of an audio bogeyman that does not always occur. But it does at least sometimes.

If a screen connects the metal cases of two mains-powered boxes it will carry some noise current. From the mains L&N via stray capacitance to the case, and from mains E directly to the case. If the screen has any resistance there will be a noise voltage between the boxes. It may be very small and insignificant in many cases, but not always. I would personally not share the return connection with a screen connection as that would add the noise voltage to the signal. If you do use coaxial cable where the screen and signal return are combined, the important thing is to make sure the screen is chunky and of very low resistance.

So my preference is for three interconnect wires. Two for signal/return and one for screen. If unbalanced, connected as per Tony's van den Hul D102/III (and many other manufacturers too) so that no screen current flows in the signal return path.

EMI does not always occur. So my personal preference will be no better than coaxial or unscreened in many cases.

 

[Tony L]

Van Damme do a cable that is available as a low, mid and high (relatively) capacitance version. During earlier trials I made up identical cables using the high and low capacitance versions of the cable (coax construction). I used the same connectors, solder etc. Differences between these cables were absolutely tiny/non existent which rather suggests that, if only using a metre, capacitance has little, or no, impact.

As ever context is everything. All you have ascertained is that there is no difference you could hear between your transistor phono stage and active transistor preamp. Change that to say a passive attenuator and a valve phono stage and you might get differing results. Use the cable as an arm-lead and you would unquestionably hear the difference!

 

[YNWOAN]

Yes transformer hum and equally I'm puzzled as to why screening would affect it. What was the larger blocker that worked?

Without wishing to derail the thread (which isn’t about mains cables etc.) can I point you to this thread in the DIY section https://www.pinkfishmedia.net/forum/threads/dc-blocker-gb-closing-soon.216403/
This is the larger blocker I used and the smaller version of the same thing is the one that had no effect. The circuit is available as just a circuit board or made up with components - you have to box it up yourself though. Pictures of the one I built are somewhere in the thread. It is worth noting that most commercially available D.C. blockers are like the smaller one I tried.

 

[Jim Audiomisc]

I'd recommend screened coax for signal connections. In part because this tends to work well in preventing external fields being picked up. Also because it helps avoid nearby bits of metal, etc, affecting the transmission of the signal.

e.g. if you use open 'twin feed' (twisted or not) then having the cable near to other metalwork, or even something like damp brickwork, will alter the loss and impedance per length of the cable. This *probably* won't matter for audio, but can be avoided by using reasonably screened co-ax.

If you have seriously long runs in a 'difficult' evvironment, then balanced screened makes sense. But I've found a need for it in any domestic audio setup I've used.

 

[YNWOAN]

As ever context is everything. All you have ascertained is that there is no difference you could hear between your transistor phono stage and active transistor preamp. Change that to say a passive attenuator and a valve phono stage and you might get differing results. Use the cable as an arm-lead and you would unquestionably hear the difference!

Yes, I quite agree. However, all the cables I have made up have been tried in the same position, between my phono and pre - so there is some continuation of context.

Despite being unable to hear any difference between the low and high capacitance VanDamme there did seem to be a difference between the shielded and unshielded I tried. Certainly not a scientific trial but it does imply that capacitance, in this instance, was not (at least solely) the cause.

 

[Tony L]

So my preference is for three interconnect wires. Two for signal/return and one for screen. If unbalanced, connected as per Tony's van den Hul D102/III (and many other manufacturers too) so that no screen current flows in the signal return path.

The Kimber PBJ (and I think most Kimber stuff) is interesting in that it is a three core open weave with one core described as a ‘drain wire’ and only connected at the source end. I assume this is their take on the source-end screen, though I’ve never understood if this is just marketing or if there is actually any engineering credibility to it. I don’t understand how part of an exposed weave could have any screening properties.

 

[YNWOAN]

So, how about this, if one is using an interconnect that uses a separate shield (so not a coaxial construction) and that shield is protecting the internal signal cables from RFI, isn’t that picked up noise being connected to the signal ground? I mean, the vast majority of these types of interconnect (often called semi or pseudo balanced) do connect the shield to the signal ground.

 

[YNWOAN]

The Kimber PBJ (and I think most Kimber stuff) is interesting in that it is a three core open weave with one core described as a ‘drain wire’ and only connected at the source end. I assume this is their take on the source-end screen, though I’ve never understood if this is just marketing or if there is actually any engineering credibility to it. I don’t understand how part of an exposed weave could have any screening properties.

Doesn’t this third wire just act as a parasitic aerial? Chord (the cable, not amp people) use a similar technique in their Tuned Aray cables - a length of extra cable connected to the return. I can’t see this doing anything other than offering the opportunity to pick up noise.

 

[Tony L]

Doesn’t this third wire just act as a parasitic aerial? Chord (the cable, not amp people) use a similar technique in their Tuned Aray cables - a length of extra cable connected to the return. I can’t see this doing anything other than offering the opportunity to pick up noise.

I’ve no idea to be honest. I’ve had the PBJ knocking about for ages and I’ve never known it to be noisy in any context. I’ve tried it with my really high-gain Leak valve amp too, and that will find hum in anything!