Cable wars 2020 - a FB experience

[CG]

18 minutes ago, barrows said:

As far as price goes, I think you guys are way off.  Take a look at the pricing for any of Belden's more complex industrial wires which use FEP insulation (not spool prices, the price from vendors who distribute it in smaller quantities).  It is pretty clear to me that the FEP insulation adds a lot to the cost (see the BAV cable interconnect pricing for comparison of a non FEP cable of the same design, I bought my GF a 15" length of BAV to use a Mike cable for recording).  And, this is not Chinese product, made by 14 year old girls in factories which do not have to operate by American standards of environmental regulations, and the labor to terminate these cables, to order, here in the USA is also high (I know what I would charge for terminating this speaker cable, with 192 24 gauge wires to strip and terminate).

BTW, I would not trust the Chinese to terminate any cable, I have seen their work.  Especially if the termination was not inspectable!

 

As for the performance, to each their own.  All I am pointing out here is that these cables were designed according to solid electrical properties and measured results, all shared by the manufacturer in detail.

 

As to whether the results are audible or not, clearly that is not a topic to be discussed in this forum space, so I will leave that to individuals to decide for themselves.

 

There is another aspect to this that rarely gets mentioned.

 

It's too long to go into detail on a forum like this, but amplifiers are inherently unstable devices.  That's because the devices used have parasitic elements associated with them and the basic processes used to fabricate the parts cause a rolloff in gain as you go higher in frequency.

 

As one example, an emitter follower stage, such as used in almost every bipolar transistor power amplifier output stage, reaches a frequency where the current gain starts falling.  This translates into a change in input impedance at that frequency.  This can and usually does cause a peaking in voltage gain.  Stability can therefore be a problem.  

 

The details of the instability and the amount depend on the devices themselves, the circuit they are used in, whatever other parasitic elements are in that circuit, and the load itself.  

 

This is why you see "Zobel" networks used at amplifier outputs in an attempt to mitigate the effects of the load.  (Bob Cordell's book on audio amplifier design discusses this at length.). There's other possible solutions, too.

 

With regard to cables, they are part of the load impedance.  That is true whether you're talking about power amplifiers with loudspeakers or preamps with the power amplifier input as the designated load.

 

So, there's an effect caused by the cables with regard to instability.  This in turn can change the characteristics of the amplifier performance.  This is mathematically demonstrable and testable - as objective as you can get.

 

Neither of these links were written by quacks.

 

https://www.eevblog.com/forum/beginners/noise-on-emitter-follower/?action=dlattach;attach=348962

 

http://audioworkshop.org/downloads/AMPLIFIERS_OSCILLATION_BJT_CIRCUITS.pdf

 

Yeah, no consumer product should have these problems.  No "pro" product should either.  How many are tested in this regard?  Have you ever seen published measurements?

 

These aren't simple and obvious details.

[CG]

Well, I'm not a guy with his own web site for these things, so I won't offer any suggestions since I wouldn't be the one doing the measurements.  That would hardly be fair on my part.  I will say that going to extremes is not necessary to show this effect.

 

My point in mentioning that one particular detail is that there's far more subtle variables than a simplified model shows or suggests.  And, these details are hardly outliers.  Or insignificant.  You can spend a a few hours with a simulation tool like LTSpice and discover that pretty easily.   (LTSpice is free, BTW.)  Or, you can measure it.  The tests usually performed are probably very accurate, but hardly even close to complete.  

 

 

[CG]

Holy cow...

 

First, if you take the time to actually do a simulation, or actually measure, you can see the peaking I mentioned.  It doesn't take really extreme cables or speaker loads to get there.  Not by a long shot.  As Barrows alluded to, you don't need to excite the amplifier into full rail-to-rail oscillations at 400 KHz to see how the response changes.  The thing is, the point the response changes too much is the point when you eventually get the system to oscillate.  Before that point isn't perfect - it just isn't to the point of sustained oscillation.

 

If you want to start small with this, just simulate the response of a very, very simple buffer stage consisting of a JFET with a JFET current source.  Vary the source impedance, just like happens with a volume control pot.

 

Again, my point is not the extreme case of smoke inducing oscillations, but instability prior to that that can and will affect sound reproduction properties.  Cordell has something to say about all this, too.

 

BTW, the situation is even more complicated when you add overall loop feedback to an amplifier.

 

But, don't believe me!  Try it yourself!  Or, just don't believe me without even trying!  I'm ok whichever way.

[CG]

45 minutes ago, jabbr said:

 

I think its natural to focus on the obvious properties of cables and the information they are transmitting.

 

I strongly suspect that many of the "SQ" differences in cables are due to the different shields and the shield effect as a wave guide channeling EMI/common mode nose from one device to another. These effects when large enough to present themselves as "ground loops" are easily understood, but I wonder when they are below the level of an obvious hum.

 

That is speculation on my part, but goes outside the obvious RLC properties of various cables (and are not modeled by LTSpice)

 

I often refer to an article by the late Pete Goudreau in regard to this:  http://www.soundstagenetwork.com/articles/pete01.htm

 

Maybe hum is the least of the problems.  Multiple high frequency signals will multiply in a non-linear amplifier - how linear are audio amplifiers at, say, 300 KHz where switching regulators might operate? - and create junk in the audio band.

 

(Comment not at all specific to or directed at you:)  It's funny how so many people just assume that audio is "easy".  I've had Fellows - Fellows as in really, really high up the food chain type with the title "Fellow" - at analog ic manufacturers say this.  The answer is always, "oh yeah - this is an effect, but nobody can hear that".  That may be true, and not for this forum.  But, doesn't it make sense to at least investigate before drawing a conclusion?

 

Here's an example:  RF guys worry about intermodulation distortion effects in connectors.  It's a real thing.  (Search terms: passive intermodulation distortion). Is that a concern in audio systems?

[CG]

2 minutes ago, Archimago said:

As a consumer, imagine if we buy an expensive laptop (amp), grab a popular USB cable (speaker cable), and what is considered a good SSD drive (speaker). We plug the components together and the devices fail to operate or potentially even cause damage!

 

Know what?  That exact thing happens all the time!  Some of the IT guys sit near me at my day job.  Every single day they deal with just that.  Precisely that.

 

Think of the drivers that don't work right.  As one example.  

 

I believe you're a medical professional, so think of the medications that end up having a bad effect on one particular patient.  Yeah, there's a long list of possible side effects provided by the pharmaceutical company, but clearly they can't test every medication on everybody under every condition.  That list hardly is encouraging to the patient, either.