speaker cable myths and facts

Status
Not open for further replies.
Reading up on this and it seems there is plenty of what some of you call 'snake oil'.

It seems low resistance is the most important criteria. Good quality 2.5mm thick cable, without silver plate looks to do the job.. The way some cables are woven is interesting and also the insulation used, but I have no idea if this is worth paying the extra money for.

For my speaker build I am looking at a few options including Shark cable which is incredibly cheap. Also considering Chord Carnival Silver Screen if it really is worth paying extra for. Wd be interested in your views and what you use in your builds whether budget or high end.
 
Well this is a touchy subject for some for sure, but I'll give my 2 cents that are just my opinion and experience..

Lets assume for an apples to apples comparison that we are using two sets of identical 14 gauge zip cord all cut to 15 feet long.. So now what most will do is put some kind of connector on both ends of the wires, which in turn get inserted into more connectors both on the amp and speaker cab etc.

If one does not have to remove the speaker cables at all or rarely, then imo soldering directly to the driver intself thus bypassing the binding posts, or to the wires inside the cab that connect to the binding posts instead, makes more sence than connecting to a hunk of metal..

Same with the amp output. If you have the nessesary skills or can pay for the service, having the speaker outputs of the amp soldered directly to the speaker cables instead of to the amps speaker cable binding posts also eliminates more hunks of metal..

Imo doing this will make just as much of a differense using the comparision stated, as upgrading to expensive over budget speaker cables..:note:
 
i did the cable thing many years ago when i was able to hear better. i bought some audioquest quartz and indigo cables (i think that's what they were called) and some turquoise and ruby interconnects (i think that's what they were called)... i sat for hours comparing day after day. i preferred the cheaper indigo's over the more expensive quartz... whatever... the differences were subtle at best. but i like the sound better when it was more rolled off on top. so some cables will do that. i heard no difference between interconnects.

when comparing to cheap, extra thin 24awg cable i thought the sound was 'smaller'... but 16awg bulk cable sounds, to me anyway, as good as the audioquest. Plus it's a lot cheaper, i can run it inside my speakers to keep things consistent, it's more flexible and i can use it liberally without worry of cost. And my system sounds excellent. I believe there are MUCH bigger factors to getting great sound. Just my opinion. Others may hear different things.
 
frugal-phile™
Joined 2001
Paid Member
Speaker cable is very system dependent. In my system i have settled on 2 24g solid strands (cryo treated, but i've not evaluated if that improves things). Certainly better matched than the typical stranded 14/16/18g twin lead.

A nice balance of performance, and frugal-phile-ism

On a dedicated woofer i'll use more strands.

dave
 
I'm with Godzilla and DaveCan on this one

20 years ago I tested a pile of speaker cables with a bunch of buddies over a weekend. We tried over and over to hear the difference as there was a lot of testosterone riding on this one. Results were not conclusive between the wazoo, 16 zip cord and 14 AWG extension cord.

Now it has spread to power cords etc...not sure why someone would not solder in a cord instead of using a detachable one but such is marketing.

The snake oil insanity was attempted to be eliminated by McIntosh 30 years ago--Audio Magazine did the same thing over 20 years ago. The problem is everyone from Best Buy, Wal-Mart to high end audio places make a fortune off the stuff. Magazines and websites make money from the extensive advertising from cables.

I can understand making money--no harm in that but when it becomes the focus of audio that is when I just walk away. I buy a steak for the steak, I won't waste $80 on a sprig of parsley garnishment no matter if it was grown in dirt mixed with the saint's ashes and watered with pee from the pope.

Audio is now like designer jeans from the late 70's. Cheap, don't work well and the pretty design matters more that what it actually does. My advice to people that ask me about speakers is this: Read about THX certification, select the level that applies to you and listen to some systems. DON'T buy one...let me know what you like. If it is OK, just buy the speakers or amps--nothing else. No stupid wires, cables, audio grade wall outlets etc.

Wires, cables, outlets, power cords etc are the Scotchguard, undercoating and floor mats of the audio industry.
 
The mathematical model of a cable is called "The Telegrapher's Equation." It is one of the first equations an electrical engineer learns that pertains to electrical circuits. It has excellent correlation with measured performance that has been tested and demonstrated for over 100 years. By manipulating the inductance, capacitance, and resistance of cables, changes to frequency response are possible depending on what the cable is connected to. It is the most expensive, least predictable, least adjustable, and all around dumbest way to change the frequency response of a sound system there is. Any other method works better. But don't tell that to audiophiles, they don't want to hear that the wire and cable industry, the one that caters to professionals and not audiophiles developed effective products and standards many decades ago for which there is little profit in a home audio system. That doesn't fit the paradigm of making profits off the ignorant.

The true test of a cable is not A versus B but A versus a shunt. There are simple ways to perform some of these kinds of tests. Interconects are easiest. Just connect an interconnect between the tape output and tape monitor input of a preamplifier, integrated amp, or receiver that is equipped with them and swtich between source and monitor. Any cable which shows a difference is defective and should be discarded. Those are invariably the high priced ones. It's more difficult to perform a comparable test on speaker cables or power cords but the cheapest 16 gage lamp cord is usually the best choice. Another inexpensive choice is 12 gage low voltage lighting cable for speakers. Do not use any cable without a UL listing for electrical power as this can put your life at risk.
 
My take on the whole thing:

Power cords- its gone through countless feet of 12-14guage solid core, butt connectors, and screw terminals, but that last 10ft from the wall makes all the difference?

Interconnects- i havent seen anything that can beat $30 canare rg59 with crimp style ends. As a bonus do you really think the pcb traces are ofc/silver shielded bla bla bla.

Speaker cable- I dont think its the cable itself. I think its the ends. Most of the banana ends ive seen are 3 pieces or more (read 3 physical pieces the signal has to go through) of brass that is press fit together and then gold plated. When you take them apart most of the time they are corroded black/green where the parts meet for the pure copper pieces.

Personally for speaker wire I would go with silver plated teflon with 2 sets of twisted pair 18 gauge just for toughness/flexibility/corrosion. You can get what, 500ft for $50.

For ends, i havent found a brand of bananas i like yet, and speakers im thinking of switching everything over to speakon.

Any tips on bananas?
 
Put me firmly in the "exotic speaker cables are snake oil" camp. :)

Only two things matter for speaker cables which are driving 4/8/16 ohm speakers - their LCR characteristics, and the reliability of the connections at each end.

(Line level phono cables have additional things to consider, like shielding against RF interference and hum pickup etc)

By far the most important is the total resistance. This is 90% of the difference in the "sound" of one cable over another right there.

Amplifiers have a certain output impedance, in modern solid state amplifiers this is very low, typically <0.2 ohms. You can calculate it based on the quoted "damping factor". For example a damping factor of 40 rated at 8 ohms is 8/40 = 0.2 ohms. Valve amplifiers typically have a much higher output impedance of a few ohms.

Speaker cable has resistance, the shorter the run or heavier the gauge the lower the resistance, and this resistance is added to the output impedance of the amplifier.

All speakers have non-flat impedance curves - usually a big peak at the mechanical resonance of the woofer, and then a gradual rise through the treble on a full range driver with no crossover.

A multi-way design with passive crossover could have just about any arbitrary impedance curve depending on the design. Some designers strive to keep the impedance curve as flat as possible (other than the unavoidable bass resonance peak) while others don't seem to give a damn what the impedance curve looks like, in which case it can really be all over the place.

It's not uncommon in some designs to see big peaks (or dips) in the impedance curve at the crossover frequencies, and for the nominal impedance to be 4 ohms in the bass but 8 ohms in the midrange in treble in a 2.5 way design.

Because the speakers impedance curve is non-flat, adding a series resistance in the form of speaker cable resistance (and amplifier output impedance) does make a real, measurable and audible change in the frequency response of a speaker. Even a half dB broadband change in the frequency response is easily audible.

In this sense speaker cables do "matter", but only because one cable is adding a different amount of resistance than another. The same effect could be achieved by actually adding a resistor. How much of a change in response adding a certain amount of resistance makes depends on the impedance curve of the speaker.

One with a fairly flat impedance curve could see relatively little change, whilst one with a largely deviating impedance curve would see significant changes in frequency response with as little as 0.5 - 1 ohms added in series.

Frequencies where the impedance is higher see less attenuation than frequencies where the impedance is lower, so you can estimate the change in response that adding resistance will make by looking at the impedance curve.

L and C properties of a cable do in theory matter, but only if you have a really long run of cable or the cable is a really unusual design. For anything up to about 5 metres I wouldn't even think about L or C, only the resistance value.

The other thing that does matter and I think sometimes gets overlooked, is contact resistance at the connectors. Any banana plug and socket, screw connector, spring loaded terminal etc has resistance. Sometimes this resistance can be quite significant, and of comparable value to the resistance of the entire length of cable.

More importantly, this significant amount of resistance occurs at a very small location - a point contact joint between two metal surfaces. Even on a banana plug the actual surface contact area is rather small due to imperfections in the fit.

Because of the small contact area the value of resistance can be unstable and can change with temperature, current flow, mechanical vibration, and oxidation of the surfaces. In extreme cases of an oxidised connector it's possible for poor contact to add distortion to the signal at high volume levels - because the current flow variations cause rapid heating and cooling of the tiny point contact surfaces altering their resistance with the ebb and flow of the music. The resistance value of the cable itself on the other hand is very stable. So if I had 0.5 ohms of total cable resistance from the wire run I wouldn't worry too much, but if the plugs and sockets were adding 0.5 ohms at their contacting faces I would be concerned.

I've had so much trouble with intermittent plugs and sockets, screw terminals, spade connectors, switches etc over the years that I now refuse to put any non-soldered connections within a speaker design except for the input terminals on the back. I never use spade connectors on the driver terminals, plugs and sockets on the crossover boards etc, I solder every joint that can possibly be soldered, as a well soldered joint doesn't exhibit the intermittent resistance variations that a connector can, especially with deterioration with age. (Both oxidation, and loss of contact tension)

Likewise if I'm making a speaker cable I always use a soldered connection from the wire to the plug - never a screw terminal connection.

So in summary, I always try to avoid connectors between amp and drivers as much as possible, soldering every joint which is feasible, choose a speaker cable gauge to keep total added series resistance below 0.5 ohms, (heavy gauge multi-strand zip cord is fine by me, as long as it meets my resistance target) making sure left and right speaker cables are the same length, even if one side doesn't need as long a run. (To keep the resistance the same for left and right, to keep channel matching the same)

The money that can be wasted on snake oil cables can be put to far better use in buying higher quality drivers and making better crossovers and cabinets - all things which will genuinely give better results.
 
The mathematical model of a cable is called "The Telegrapher's Equation."
No.
From: Telegrapher's equations - Wikipedia, the free encyclopedia
The telegrapher's equations (or just telegraph equations) are a pair of linear differential equations which describe the voltage and current on an electrical transmission line with distance and time.
So whats a transmission line?
a transmission line is a specialized cable designed to carry alternating current of radio frequency, that is, currents with a frequency high enough that its wave nature must be taken into account.
Audio frequencies (and far higher frequencies) combined with speakercable lenths don´t allow currents to be waves.
 
The lumped sum equivalent of the distributed parameter network that describes wires works perfectly for predicting cable behavior in electronic components and between them. There is no reason to doubt that the same effect created by strangely constructed wire intended to produce a given frequency filtering effect can be achieved far more predictably and cheaply through other means such as constructing an equivalent passive filter network or using an active equalizer. The advantages of the active equalizer are numerous including great flexibility, low cost, and the fact that because the filters are between buffer amplifiers their action is entirely predictable, largely independent of the source and load impedence. Wire used as a filter network has none of these advantages.

A $1 RS 3 foot interconnect carries a 7mhz NTSC detected video signal that is indistinguishable on screen from the same signal injected as RF into tuner of the best 36" analog TV set of 11 years ago. That's 350 times the bandwidth of an audio signal and of course the response is flat down to DC. This cable has no performance flaws. Its one drawback is that it generates no significant profits for anyone either.
 
Hi,

Reading up on this and it seems there is plenty of what some of you call 'snake oil'.

Yes, there is.

It seems low resistance is the most important criteria.

Now there is some major snake oil right here. Anyone claiming that low resistance (beyond a point, say no more than 10% of the total DCR of bass drivers voice coil and crossover choke - meaning usually anything under 0.5 Ohm) is clearly marketing snakeoil, unsupported and in reality contrary to facts.

Good quality 2.5mm thick cable, without silver plate looks to do the job.

If that is what you believe, sure, go for it. It may be worth to compare this to the even more incredibly cheap 1mm solid core mains ring wire (parallel the bare earth conductor with black for "-", the other is "+"), just for laughs and giggles and to see if my "low resistance in speaker cables is snakeoil" quip actually has some foundation in reality.

Wd be interested in your views and what you use in your builds whether budget or high end.

Well, in my system I use Tempflex SCSI III ribbon cable, this has 80 very thin solid core copper conductors, silverplated, with PTFE insulation. I am usually using solid silver (goldplated or enameled) for internal wiring in speakers (and all other wiring, except mains).

The Tempflex cable is not really "Audio" cable and it is not cheap (around 5 USD per foot last time I bought some).

Ciao T
 
Hi,

Power cords- its gone through countless feet of 12-14guage solid core, butt connectors, and screw terminals, but that last 10ft from the wall makes all the difference?

Well, anyone who has passed EE101 SHOULD be able to explain why it is precisely the the last 10 Foot that make the difference and exactly what processes contribute and how we can minimise the sensitivity of equipment to these effects and how one may construct power cable so as to significantly improve the situation with "legacy" equipment not build by applying such knowledge.

Interconnects- i havent seen anything that can beat $30 canare rg59 with crimp style ends. As a bonus do you really think the pcb traces are ofc/silver shielded bla bla bla.

So you judge "beating" by looking at cables?

For fun, you can compare Canare RG-59 and a higher grade double shielded (and thus having a much lower return path impedance) RG-Series cable that has essentially identical electrical and mechanical parameters otherwise. And you can apply correctly made soldered ground connection to that crimping and use a plug that can be made to have a very low ground resistance by allowing the plug ground collet to be tightened (e.g. older WBT style).

For extra credits, using only knowledge present in EE101, please explain why the cable with the significantly lower ground impedance end to end would be perferrable in this application and what the mechanisms involved are.

Finally, in better quality gear especially since the passing of ROHS in europe the PCB traces are often goldplated (as tin/lead plating is not compliant).

Speaker cable- I dont think its the cable itself. I think its the ends. Most of the banana ends ive seen are 3 pieces or more (read 3 physical pieces the signal has to go through) of brass that is press fit together and then gold plated. When you take them apart most of the time they are corroded black/green where the parts meet for the pure copper pieces.

A rare bit of common sense here.

I personally prefer single piece spades, copper, not Brass. These used to be available from Radioshack in the UK cheaply, when they closed all shops I bought all the available stock in London and nearby.

Spades and good quality single piece binding posts can make a very low resistance connection and if regulary checked for tightness the results are very long term reliable.

Personally for speaker wire I would go with silver plated teflon with 2 sets of twisted pair 18 gauge just for toughness/flexibility/corrosion. You can get what, 500ft for $50.

I would instead suggest that if cost is a major concern, plenum rated Cat5/5e/6 etc Network cable (the solid core version used for the main wiring) can be used (make sure the conductors are copper). It is normally 8 pcs of 24 Gauge solid copper in PE or PTFE/FEP insulation.

Many years ago I published notes on the FFRC Speaker cable on TNT-Audio. These writings and subsequent articles introducing more complex constructions by others may be a good idea to read.

Such wire is very cheap, I regularly used to blag partially used boxes from network contractors for ought... Even paying for 300m Reel will not bankrupt anyone.

Any tips on bananas?

Yes, Eat them if you like them, but keep them away from speaker cables.

Ciao T
 
Hi,

Amplifiers have a certain output impedance, in modern solid state amplifiers this is very low, typically <0.2 ohms. You can calculate it based on the quoted "damping factor". For example a damping factor of 40 rated at 8 ohms is 8/40 = 0.2 ohms. Valve amplifiers typically have a much higher output impedance of a few ohms.

All this is meaningless. Any REAL calculation of the "damping factor" must include the speakers voice coil DCR and the resistance of any inductors.

In which case we see that real damping factor quickly approach unity, as long as we take the rule "rated impedance must not be greater speakers minimum impedance plus 30%".

In other words, the whole talk about damping factor is one big red herring or bottle of snake oil.

Sure, there is a sensible limit as how much series resistance will be too much, however, given that moving ones head a few inches easily causes frequency response changes of many dB, even 1-2dB change in frequency response is in practice unlikely to be meaningful in the sense of degrading performance.

Ciao T
 
Hi,

I've also seen those green insides of corroded banana plugs-Are Neutrik connectors a much better way to go?

The Neutrik Speakon connector is a MASSIVELY BETTER alternative to anything in use in High End HiFi, but it has little to no compatibility with general HiFi gear.

If you do not need "normal HiFi compatibility" make a straight bee-line to Speakon and consider using them bi-wired as well...

Ciao T
 
Hi,


Because they are designed for line level balanced signals, not for speaker connections.

For speaker connections the correct connector is called Speakon and Neutrik has also an alternative, much improved power connector to replace IEC, called Powercon.

I can strongly recommend both of these.

For Balanced Line connections I prefer all plastic Cliff brand XLR's.

Ciao T
 
Status
Not open for further replies.