Long Speaker cable, short I/C, or Short speaker cable, long I/C?

Alan

As above gents, is it better - in your view - to keep short interconnect cables and have longer speaker cables?

Or to have long interconnects (with mono amps obviously) and short speaker cables?

Interested in your experiences, views, unqualified opinions, thread craps, puns, bad jokes and anything in between. Also your opinion Col, whats the bst way to use yer TQ?

uglymusic

Creak! Hiss! (that's the sound of a can of worms being opened :-) ).

Docfoster

I have seen monos put on the floor next to speakers, from which I infer that some people think short speaker cables are the priority. Or may be that's keeping power amps away from delicate sources. I don't know.
As ICs are more expensive per unit measurement, I've always been one who has kept the amps close to the sources.

Alan

I gather the same Ben, and I/C cables are more expensive. But...Some strongly prefer the short speaker cables.

It occurs to me that a low level signal such as is transmitted through an I/C would be vulnerable, fragile even, easily damaged, compared to an amplified signal with a grunty amp pushing it down a speaker wire.

uglymusic

My speaker cables are well more expensive than my interconnects, so that's not always the case. The situation gets even hazier when you're multi-amping.

Alan

I appreciate there is a strong economic factor in this case, but I guess I mean simply in terms of subjective preference and/or technical explanation.

My source is about 4 metres away from the left speaker in terms of speaker cable length, and about three from t'other.

Jim

I have always preferred short interconnects and longer speaker cables.  A preamp, whether active or passive, has a harder time (doesn't mean it can't be done, just that it's more difficult) driving longer interconnects than a power amp does driving speaker cables.  Further, the signal levels are lower and it's easier to bugger them up than it is with what comes out of your power amp.

Alan

I understand what you're saying Jim, in fact you have expressed something like what I meant to suggest in a much clearer way.

Are speaker cables generally shielded? I see a lot of I/Cs are.

uglymusic

I used to run a tri-wired set-up with short AudioQuest cables - thick, expensive buggers - and long interconnects. That always sounded good, but I admit I didn't try long speaker cables and short interconnects on that rig.

Jim

I understand what you're saying Jim, in fact you have expressed something like what I meant to suggest in a much clearer way.

Are speaker cables generally shielded? I see a lot of I/Cs are.

There have been some screened speaker cables in the past, I believe Townshend did some.  I think there not generally available for two reasons: first they would add even more to the cost and second there's no need in that the power levels are higher and its susceptibility to noise is lessened.

No reason why they can't be though.

uglymusic

Col's our speaker cable man (amongst other things), of course.

PAC

Short speaker cables are preferable unless you have an NVA/Linn/Naim amp that requires a minimum inductance loading (usually needing a minimum of 3m or so per side to generate the inductive load).

Signal cables are low current low voltage and are usually connected to high impedance (47KOhm) loads so need very little power to drive them large distances.  You can get away with 10m signal cables without loosing much at all and providing they are low capacitance and well screened, there's usually no problems with runs of this length except for from TT's to phono stages where the capacitative loading would be undesirable. The one exception is where you are connecting a relatively high output impedance source to a relatively low input impedance amp (where the ration of input to output impedance is less than 10 times).  In this scenario, the cable capacitance and impedance can result in a low pass filter being formed in the circuit which can roll off HF response.

long speaker cable runs don't forget add to the reactive circuit impedance of the amp/speaker circuit and to the total capacitative load which may be undesirable due to the effects on the amplifiers output stage (can cause ringing in the output circuit).

PAC

Ps....avoid screened speaker cables...they're not needed for short/medium runs and add considerably to cable capacitance.

uglymusic

Oh, and Paul's our cable man, as well. Oops!

Alan

Thanks Paul, very interesting. This is way above my understanding of course, but I have never, ever seen a requirement for an NVA amp to have a minimum cable length, nor have I seen ideal inductance specified - only the need for a) low capacitance and b) no litz braid or goertz construction.

I may be wrong, or not, but it seemed worth mentioning.

Brumjam

Very interesting Paul.

PAC

I quoted NVA/Linn/Naim as they employ very little (and in some cases none) inductance in the output stage.  NVA may not quote a figure so may have a minimal amount of inductance but some Naim amps don't use inductor coils in the output stages...instead they rely on cable inductance to present enough resistance to prevent instability (oscillation) in the output capacitor stages.  Think of it as a buffer or damper without which the amp can go into undamped overload.  Speaker cables have inductance, and the reason naim and others rely on the cable is that their amp design relies upon this to make their exact circuit specificatins for best performance.  Indeed, if a contemporary twisted pair cable with say a capacitance of 80pF/m or more were used with Naim amps, it could destroy the output stage.  Think of an analogy where the circuit is all on one board with your speaker loading and output at one end and your amp output stage at the other.  In between you need to see a certain inductance rating.  If you swap out the inductors for say a large capaciitor, you could fry the output stage.  The speaker cable acts as a capacitor, and the longer the length, the greater the value of the capacitance and the greater the reactive load.

Thats why if running Naim amps, there's no better cable than Naim's own speaker cable and at £16/m its not that expensive.  It introduces controlled inductance and completes the required inductance circuit between amp and loudspeaker for optimum performcance.

Low capacitance and higher inductance are not a problem with modest speaker cable runs for more contemporary amp design.  Its not until you exceed 10 or 20m lengths that the high inductance can start casuing audible HF roll off, but if the capacitance is high, it can cause ringing (sounds like harsh clipping). That's why I'd always advocate a low capacitance speaker cable design of minimal impedance.

Impedance is the other consideration.  the higher the impedance the greater the total impedance load.  If you have say a 10m run and the cable measures say 0.2Ohms per metre (a bit high but will serve as an example).  If say the speaker presented a nominal 8 Ohm load, then the total load would become 8 Ohms plus (10m x 2 x 0.2) = 4Ohm cable loading which is dissapated as heat = wasted energy and additional load on the amp.  Total direct load becomes 4 Ohms plus 8 Ohms = 12 Ohms.  It doesn't end there.  As well as nominal impedance, the cable has a reactive impedance which varies with frequency, based on capacitative and impedance load with frequency (as frequency rises, so does impedance).  I haven't yet mentioned phse envelope, but combine a peaky phase response with a high cable loading and your amp is really having to work hard!

My two rules of thumb with speaker cables:  For shorter runs of a few metres or less, you're unlikely to hear a huge difference between many speaker cable makes irrespective of cost providing they exhibit similar LCR characteristics.  As the runs increase, then aiming for low impedance and low capacitance is a must for best performance.  You should generally be safe if using a 2mm2 OFC section for most domestic runs, but avoid any fancy highly capacitative designs such as multiple litz braid for example.  Its a poor excuse for a design relative to speaker cables.  You may benefit by more exotic geometry such as true litz (not braided litz) as skin effect can have a minor impact at longer domestic runs (say 10m or so) as does phase shift, but in reality, upgradig say a cartridge may have a more pronounced effect!  All you can do is experiment in your own system.  As for me, I've ditched my Chord Odyssey cable in favour of Talk3.  Its kinder to the amp and I reckon sounds better even though it costs 5 times less!

Alan

OK mate, I and others no doubt will appreciate the time you have taken to expain that.

Brain_Dead

Quote PAC "
I quoted NVA/Linn/Naim as they employ very little (and in some cases none) inductance in the output stage"

Having worked with Naim, they do have a Snubber/Zobel Networks and often have low value wire wound resistor to add inductors, now in test I did for Naim on the Nait 5i and the Super Nait , the 5i would drive full power for hours in 10uf on a short cable the Super Nait that was wicked 100uf was added to the terminal causing large amounts of heat but the bugger never died, has to inductive loads both were very good driving 1KVA toroidal on shout runs of cable, a well designed, well engineered piece of kit.

Now RD says the NVA are not designed so well as the Naits because he has forgotten to fit he Zobels OK a design or cost choice. the Linn are now mostly Class D I thing and there is a huge lump of magnetic filter inductor in that design.

And to assume on a unknown length of cable and it track around radiator of steel is not a good design regime if you want you kit never to come back for repair. 

~O) ~O)