[Karsten]

I guess most has experienced that feeling of highs that sounds like it is about to tear your ears apart, it just sounds terrible…. Well, new preamp, amplifier, cd player, cables and so on may seem to solve that problem, voila the inferior component has been located, or?

An interesting observation I have done, which seems to be pretty consistent over the years, is that better speakers allows for a more transparent front end.

I had one of these experiences not so long ago. My SP Revelations were using the previous prototype cross over, which also sounded pretty good. A guy came by with a modified SlimDevices Transporter which he would like to try in my system. I was amazed because it sounded substantially better in my system than the PC based front end I’m using. On the album Eric Clapton & BB King, there was no doubt that Claptons voice and the guitars sounded a lot better with no harshness at all. I was pretty much prepared to write off the PC front end, because it sounded “digital” and messed up the highs on some albums. I always thought is was the recordings that were not good enough, but with this new “evidence” I pretty much had the proof that that was not the case.

Then I did the upgrade with the top-of-the-line Mundorf components, which gave an immediate and substantial improvement. I called the guy and told him about it, because as good as it was sounding now with my PC based front end, I was curious about how fantastic it would sound with this magic Transporter. He came and left the Transporter with me a couple of days. It quickly became obvious that the magic was gone… I connected both the Transporter and PC to my DAC and did some A/B comparison. Now it was very clear what the Transporter did compared to the PC, it was simply slightly fleshing out the sound, slowing down all those little fast transients a bit.

The pattern I see here would indicate that a lot of the effort I have been using in the past on optimizing my system, basically has had the goal to slow down the transients just enough in order not to exceed the limitations of the speakers, because when your front end exceeds the capability of the speakers, it will translate into bright or harsh sound…… Hmmm, somehow this makes sense to me and explains a lot of issues I have had in the past with component matching etc.

Now, after this latest cross over upgrade, I just need the cleanest front end I can get and it translates into the sweetest and most accurate sound I have ever heard.

Karsten

[Aether Audio]

Friends,

In response to Karsten’s observations, I feel compelled to offer some form of “hypothetical” explanation as to what is going on here.  As we all know, the area of crossover components, wire and interconnects is a hotly debated issue in audio.  I’m not pretending to offer concrete explanations – just food for thought.  So please take the following for what it’s worth.

According to Karsten, the observation is that when fast rising transient signals are fed into the speakers with (supposedly) inferior crossover components, a type of high frequency “glare,” “artificialness” and irritating sound results.  This is very interesting from an engineering standpoint.  It would seem to imply that distortion products are being generated somewhere in the signal chain – most likely in the crossover itself.

Another interesting fact here is the observance Karsten  has made with regards to the “Transporter.”  This suggests that if a device is placed in the signal path that band-limits (low-pass filter function) the source signal, i.e., reduces its rise-time/slew rate (speed), then the offensive condition is reduced and/or eliminated – albeit at the expense of overall transparency and resolution.  This is potentially a powerful clue as to what is going on.

In order to break things down to a manageable level, I am going to categorize what I believe to be the three major sources of potential distortion.  Listed in order of significance, they are as follows:

1) Microphonic Vibrations

It is a well-known physical fact that electro-strictive and magneto-strictive forces can arise in electronic components.  The analysis of these effects is rather involved so I am leaving it up to the reader to investigate further if they so desire.  Suffice to say, electrical components can and usually do mechanically “vibrate” at some certain range of frequencies if sufficient voltage is applied across and/or current through their terminals.  Most are familiar with the conspicuous “hum” produced by a 60Hz power transformer. 

Once such a device is “excited” into mechanical vibration, the conversion of electrical energy into mechanical energy can (and does) reverse.  The mechanical vibrations will in turn cause electrical signals – albeit orders of magnitude smaller in amplitude to that of the electrical stimulus – to be produced.  In other words, a component that has been excited into mechanical vibration will then produce electrical signals of its own.  These can be either clearly audible or scarcely detectable, depending on their magnitude.  If you don’t believe me, just rap on your preamplifier or amplifier chassis while no music is playing and the volume is turned up.  You’ll likely hear some analog of that action emanating from your loudspeakers – but be careful!

How does this relate to transients?  A fast rising transient is akin to a drumstick striking a drum.  Since the mass of the drumstick does not change, the only way to achieve a louder “snap” of the drum is to increase the velocity of the stick as it approaches the drum head.  The increased velocity of the stick hitting the head is translated to greater energy being transferred from the stick to the head.  This results in not only an increase in the resulting amplitude, but also an increase in the number of harmonics generated by the head.  That’s why the timbre of a drum changes depending on how loudly it is played.  The increased rise-time of the initial transient contains more energy and therefore stimulates the production of more over-tones. 

This, I believe, is the same mechanism at work in the observed phenomena of Karsten’s observations.  Any device placed between the signal source and the speaker/crossover system that “slows down” the rise-time of the source’s transient content, will be effectively the same as slowing down the drummer’s hand/drumstick in the example above.  Less energy = less mechanical vibration = less harmonic excitation = less distortion.  Simple enough.

Superior crossover components are constructed to more rigorous mechanical standards and are therefore more mechanically stable.  Theoretically, they should suffer less from the electro and magneto – strictive effects outlined above.  Both inductors and coils can suffer from these effects and there are methods that can be employed to minimize them.  Inductor wire can be wound more tightly, encapsulated in varnish or epoxy or made of flat ribbon instead of round wire.  The flat ribbon material was chosen for inductors used in Karsten’s latest “ultimate” crossover construction.  Also, Capacitors can be wound tighter and constructed more rigidly as well.  Mundorf is well known for the level of rigor they apply in the construction of all their products.  Karsten also chose to build his crossovers externally from the loudspeaker enclosure and its unavoidable, low-level vibrations.  I suspect this played a significant role as well.

[Aether Audio]

2) Hysteresis Effects

It is well known that all reactive components (inductors and capacitors) carry with them the potential to exhibit hysteresis effects.  I will not attempt an explanation of hysteresis, the reader is free to research this topic on his own as well.  In the previous “Mundorf” thread I discussed the effect and Karsten’s discovery of improved performance resulting from the upgraded inductor in the woofer circuit.  Since magnetic materials are widely known for exhibiting this problem and we have already covered it, I will not beat that dead horse any further.

What is not often realized is that capacitors can exhibit hysteresis as well.  Instead of the common “B-H loop” referred to in magnetics, with capacitors we have the analog “D-E loop.”  Such non-linearities arise from a number of related factors, most all of which have to do with the dielectrics used in their construction.  Dielectric absorption (DA) is the most notorious offender, yet in the link provided below the author discovered that “paper & oil” capacitors seemed to be the most linear.  P&O caps are notorious for having very high levels of DA.  Clearly, there are mysteries here that have yet to be scientifically dissected.  Nevertheless, it is a well known fact that hysteresis in any form generates distortion.  Less distortion is good.

http://members.aol.com/sbench102/caps.html

So how do fast rise-times affect this?  Well, remember…fast rising signals contain more energy.  Also remember that the voltage across a capacitor lags behind the current through it by 90 degrees.  This means that when the current is greatest (i.e., in a sinusoidal waveform), the driving voltage producing the current flow is at zero.  This means the alignment of the molecules in the dielectric must achieve their polar orientation during a period when there is the least amount of electrostatic force compelling them to move.  Any tendency for “sluggishness” in their movement will result in some level of hysteresis being exhibited.  The faster they are required to move, the more such hysteresis may become evident.  Superior dielectric materials will reduce these effects and would explain their reputation for superior sound. 

It is interesting to note that lower voltage signals will produce less of the electrostatic force that compels the dielectric’s molecules to move.  A combination of low voltage, fast rising signals may represent the “worst case scenario” for a dielectric source of hysteresis.  As in Karsten’s case, the claim of superior performance resulting from higher quality capacitors is in the area of micro-dynamics and harmonic detail.  Both of these are the result of low voltage and high frequency signal conditions.


 

[Aether Audio]

Burn-In Demystified

In fact, I now believe dielectric behavior is at the root of the “burn-in” issue.  Even cables are made of insulating materials that effectively constitute a dielectric.  I suspect that when these materials are “green” the molecular bond between adjacent molecules of the dielectric are relatively strong.  Hence, each molecule is less free to change its orientation in response to the applied electrostatic field, when in initial operation.  After being subjected to an alternating current for some period of time, these initial bonds “loosen” and the molecules then respond more quickly to the applied signal.  This in turn, reduces hysteresis artifacts and lowers any resulting distortion. 

Please don’t misunderstand…the issue is more complex than this simple analogy.  The molecules in the dielectric aren’t free to just go wandering around after burn-in.  The point is that this is (or may be) “sort of” what’s going on.  Any further explanation requires more understanding than I possess.  The upshot though it that’s why I believe (at this point in time) your cables and everything else sounds better after some period of burn-in.

[Aether Audio]

3) Conductor Properties

This is an area of much mystery and the least obvious source of any distortion.  The behavior of conductors and the various alloys used in their construction has been thoroughly investigated over many years and the jury is still out.  Copper vs. silver vs. Litz construction vs. solid core goes on ad infinitum.  I think I have some vague clue as to what’s going on but it is exceedingly complex.  I believe that at the root of it, the issue of electron migration through a conductor and the resulting “electron cloud” of charge carriers within it is at play.  I suspect it may even relate to atomic nuclear alignment and it’s effect on charge carrier migration along with quantum mechanical wave motion within the conductor.  Whew!   

Something tells me that the actual charge carrier vector and any disassociation (always present in real conductors – hence “resistance”) between it and the ultimate Poynting energy vector (the loudspeaker load) yields spurious signals that are not harmonically related to the music.  The claimed “diode effect” of two or more un-insulated strands of wire contacting one another along the length and in the construction of a larger conductor, would tend to exhibit a “worse case scenario” for this effect.  The end result could be envisioned as a form of “noise” clouding and confusing the reproduction. 

Also, various alloys and doping of conductors may impart a “diffusing” effect upon these small artifacts.  These elemental “impurities” may tend to “diffuse” extremely small standing waves (very short “quantum level” distances) that develop orthogonal to the main Poynting vector.  Such an effect would go a long way in explaining the differences claimed between copper and silver wire.  Copper may inherently possess a greater “diffusing” effect due to its more random charge carrier movement (higher electrical “resistance” than silver) in the crystalline grain structure of the conductor.  One variant that is commercially available that may be addressing this whole problem is Stealth Audio’s “Indra” interconnect.  It is claimed to be constructed of an amorphous “metal” or “metallic glass.”  As such, it would have no crystalline structure whatsoever and therefore exhibit completely randomized charge carrier movement.  Such random electron motion would (theoretically) prevent the formation of these quantum level standing waves.

In the case of Mundorf’s “silver-gold” alloy, I suspect the few gold atoms interspersed in the silver act as such a diffusing mechanism.  The advantage of this combination is that the superiority of silver’s higher conductivity could be realized while avoiding the quantum level effects of orthogonal charge motion.  The gold atoms and their significantly higher mass would be immovable “barriers,” blocking the path of these quantum waves and diffusing them - not unlike various diffusers used for acoustical room treatments diffuse standing waves therein as well.

If such phenomena are actually being manifest, then that begs the question once again; “How do fast rise-time signals aggravate this problem?”  Well, fast rising signals (voltages) translate to faster charge carrier motion (current).  If orthogonal currents within the conductor are forced to move faster, then they are more likely to stimulate these quantum-level standing waves.  Think of the antithesis:  DC current flows in one direction only.  Electrons don’t change direction in a back-and-forth motion as in AC current (music signal).  Such an unchanging current flow would not be expected to induce any such “standing waves.”  What is required is an “impulse” to get them going.  The stronger and faster the initial impulse…the stronger the resulting standing waves.  This is no different than one clapping their hands in a large auditorium.  The harder the clap…the louder and longer the resulting room reverberation.  The only question is…is that what’s really going on?  It makes for an interesting “guess” though, doesn’t it?

Obviously and in any case, these artifacts  - if they exist at all - are quite small in magnitude compared to other sources of distortion and will only manifest themselves in the most revealing of systems.  Thankfully and whatever the causes may be, Karsten has shown that our “core technology” here at SP Tech provides such a platform to evaluate these very small effects – at least as far as we’re concerned.   

Hope this doesn’t confuse you more than it helps.

-Bob

[Karsten]

Bob,

Thanks a lot for your hardcore theories, interesting reading.

In this case I don't think that microphonics is the major cause, since I had the other prototype cross over networks externally as well.

This theory seems to be the most plausible:

"It is interesting to note that lower voltage signals will produce less of the electrostatic force that compels the dielectric’s molecules to move.  A combination of low voltage, fast rising signals may represent the “worst case scenario” for a dielectric source of hysteresis.  As in Karsten’s case, the claim of superior performance resulting from higher quality capacitors is in the area of micro-dynamics and harmonic detail.  Both of these are the result of low voltage and high frequency signal conditions."

The level of detail I'm getting is almost scaring high, on several occasions I found myself thinking "Well, if the mastering engineer had been using these speakers, he/she would have put some more attention to this or that matter"

I had a friend over yesterday. In the past he was very much into electrostatic speakers, his claim was that these Revelations combined the best parts of the electrostatics with the best parts of dynamic speakers and frankly I agree with him.

We were also listening to the Sinead O'Conner "Am I Not Your Girl" which is a pretty good recording and I have actually used it a reference in the past. What we both noticed was that now this was a dead give away as a studio recording. She is very obviously not in the same room as the band but standing in one of those small recording cabins and then some reverb has been added to her voice.

Karsten

[WOR Radio]

Never discount the notion that the mastering of the recording could be inferior.

In my book..NEVER NEVER NEVER, rely on one or a few recordings when voicing loudspeakers. It takes a vast cross-section of a wide array of recordings to reach the happy medium after your scientific testing is accomplished.

Just because you have a "known recording"..do you really know it unless you had a hand in its production?

See?

[Karsten]

Never discount the notion that the mastering of the recording could be inferior.

In my book..NEVER NEVER NEVER, rely on one or a few recordings when voicing loudspeakers. It takes a vast cross-section of a wide array of recordings to reach the happy medium after your scientific testing is accomplished.

Just because you have a "known recording"..do you really know it unless you had a hand in its production?

See?

You can be sure that I'm not relying on one recording, just mentioning some findings and exambles. Colorations etc. are not that difficult to spot if a good amount of recordings are used. Right now I'm just trying to max out the resolution and get any signature from upstream electronics out of the way, at some point I may "back off" a little and throw in a different speaker cable, preamp or something like this.

Regards,
Karsten