Speakings of Signal Width

reeltrouble1

So after the question of power and watts and current pushing electrons down a path it would seem logical to talk about the actual path. Afterall, if all things were equal then things just would sound the same.

So what about this path? Well, we pretty much can agree that the path should offer low restriction, nice to say but what does this really mean, we like to say keep the path short, but a short path would not necessarily be low in restriction.

Its just a fact of audio that any gain stage either active or passive will restrict the flow of the signal. This is a bad thing and does effect what you hear, be it soundstage, dynamics, bass, and so on. So manufacture's of high quality gear will typically keep the gain stages to a minimum.

So what about the concept Signal Path Width? After some thought it seemed almost obvious that a path can be short in stages but if it is too narrow the signal will be restricted and display poor sonic qualities. I have found Victor Khomenko's thoughts about this topic interesting. He identifies this signal path width as quiescent power and places great importance on it as a predictor of good sound. Only one part of a puzzle, but still quite possibly vital. As an example a Class B amplifier may be very powerful, but with a low quiescent power by its design it would not sound that great.

Of those who care about such things how say you?

RT1

hearingimpared

I'm a little confused Teddy. Does signal path width have anything to do with bandwidth?

heiney9

One of the reasons I like Nelson Pass's philosophy is he believes in very few, usually 3 or less, gain stages in his products. He also believes in using the best parts for the intended price point, extreme matching, and circuit designs that don;t need extraneous parts and circuits to help them perform and behave as intended (ie; negative feedback, complex current limiting, etc)

His designs always seem to get rave reviews so he and RT1 maybe on to something.

It seems logical to keep the paths the signal travels unimpeded and as short as possible along with components/topologies that don't restrict or require add'l parts/circuits to enable those parts to perform as intended.

One thing is absolute in audio and that's the concept of compromise. A good designer understands that there are compromises everywhere in even the best designed equipment. It's limiting the compromises to those we can live with and are beneficial to sound reproduction. This can be achieved in various ways and as always some designers are willing to compromise more than others in certain areas.

Cables make a HUGE difference in the overall synergy and enjoyment of my rig and that comes from the concept Ted mentions concerning the flow of the signal.

H9

ShinAce

reeltrouble1 wrote: »

As an example a Class B amplifier may be very powerful, but with a low quiescent power by its design it would not sound that great.

Of those who care about such things how say you?

RT1

I don't get your example. Are you arguing that amps with higher idle current sound better because of topology and design layout?

heiney9

ShinAce wrote: »

I don't get your example. Are you arguing that amps with higher idle current sound better because of topology and design layout?

Read this it might enlighten.

http://www.passlabs.com/pdf/articles/distortion_and_feedback.pdf

Another superb article and it parallels what we are discussing here. If you make it thru to the end you'll have the answer to your question.

H9

ShinAce

I read the article, and I still don't understand if we're actually on the same page.

Heiney, you yourself said earlier that minimizing components is a good philosophy, and even included parts not in the signal path:"It seems logical to keep the paths the signal travels unimpeded and as short as possible along with components/topologies that don't restrict or require add'l parts/circuits to enable those parts to perform as intended."

I'm sorry, but even if I'm building a single stage class A amp, the most important aspect, short of the power transistor, is the current source. I could minimize part count by using a wire-wound resistor, but who does that? Don't we all end up going to complex and tiered current sources?

Of course for the same parts, a shorter path is better, to a certain point. If I have to put sharp bends in the traces just to shave fractions of a mm, I'm not doing it.

All that to say; what is signal path width? All other things being equal, what one variable would change the 'width' of the signal the most?

heiney9

Yes, I did say that but you need certain basic components. Just like a car, the lightest, most efficient car, might just include the wheels and tires and nothing else. That doesn;t really make a car.

I don;t know every single commercial design but I know of no single gain stage amps that are practical.

The article I linked to had to do with the part you quoted from RT1 about class B and idle current. Probably not directly on point but within the fringe. Plus it's a damn good article.

reeltrouble1

hearingimpared wrote: »

I'm a little confused Teddy. Does signal path width have anything to do with bandwidth?

I would think they are related, however, when Khomenko is speaking of signal path width he is describing a measurement of the stage's width.

I have linked the white paper and the section I am referring to is covered under the paragraph's entitiled Low Restriction Signal Path.

http://www.balanced.com/resources/whitepapers/VK-60white.pdf

BigMac

What are your thoughts on something like this?

http://www.audiophonics.com/audiophonics-zeus-about.html

reeltrouble1

ShinAce wrote: »

I don't get your example. Are you arguing that amps with higher idle current sound better because of topology and design layout?

Well I am not arguing for any particular topology, Khomenko suggests that the width of the stage is just as or even more important than the number of stages. Now I have owned several of his amps, so my empirical data from listening experience tells me something special is going on with the gear, his white papers talk about many facets of component design, however, I found this concept of quiescent power as something different than other things I have read.

Just throwing it out there as I cannot recall it ever being discussed.

RT1

treitz3

Jimmy Crack Corn and I don't care. All I care about is the end result as to what hits my ears, therefore I'm out of this discussion.

Rock on......

dkg999

My head hurts now

Does Amar Bose have any papers on this issue?

I'm kind of a fan of the magic of audio, watching an LP spin around, amazed that sound even comes out of my speakers, etc. Not sure I need all these details!

BigMac

BigMac wrote: »

What are your thoughts on something like this?

http://www.audiophonics.com/audiophonics-zeus-about.html

Wondering if this is what you are talking about in terms of the end product. Less is more, gain or no gain??. Just trying for it to make sense to me all the way and to further my knowledge on this topic.

ShinAce

reeltrouble1 wrote: »

Well I am not arguing for any particular topology, Khomenko suggests that the width of the stage is just as or even more important than the number of stages. Now I have owned several of his amps, so my empirical data from listening experience tells me something special is going on with the gear, his white papers talk about many facets of component design, however, I found this concept of quiescent power as something different than other things I have read.

Just throwing it out there as I cannot recall it ever being discussed.

RT1

I'd be interested in talking about quiescent current exclusively before adding in these extras. With a firm grasp of the idle current, we can move on to more detailed discussions.

The white paper you linked defines the signal stage width as a measure of its quiescent current(I say current because of the amplifier types we will discuss). Quiescent power may be more appropriate, but we don't need to refer to it.

heiney9

There is some great reading here about amplifier topologoies, etc. Granted it;s from a single persons POV. The "Leaving Class A" is another great article.

http://www.passlabs.com/articles.htm

Not sure what you mean by quiescent current exclusively. Quiescent current comes into play based on the topology. The overall topology is much more important than the role the quiescent current plays singularly.

I'd pick apart and question the topology before I'd worry about anything related to the quiescent current in the various stages.

Quiescent current is idle current which is the main topic in the "Leaving Class A" article.

ShinAce

heiney9 wrote: »

Not sure what you mean by quiescent current exclusively. Quiescent current comes into play based on the topology. The overall topology is much more important than the role the quiescent current plays singularly.

I'd pick apart and question the topology before I'd worry about anything related to the quiescent current in the various stages.

Quiescent current is idle current which is the main topic in the "Leaving Class A" article.

How does the idle current come into play based on topology? Usually, when we talk about topology, we're talking about 2 stage versus 3 or 4 stage amplifiers using different combinations of current, transconductance, and voltage amplifiers. This has nothing to do with the idle current.

Quiescent current is idle current by definition. This is set by the bias of the amplifier. If we take an AB amp, we can bias it into class A operation, or shortchange the bias into class B operation. The topology is the same but the class has changed simply because of bias.

I still don't think we're talking(or seeing) the same thing.

W WALDECKER

Victor Khomenko is a brilliant Audio designer , physicist and electrical engineer. First and foremost he is a audiophile .He designs some beautiful sounding gear and i am a big fan of his and i admire his work very much. I also read his white papers that are on the BAT website and they are also included inside of the owners manuals.thanks....WCW III

mhardy6647

Wanna learn the basics of hifi (EDIT: analog) electronics (which haven't changed since the 1930s)? Read these three short, well-written books, available on-line courtesy of the great Pete Millett:
http://www.pmillett.com/Books/crowhurst_basic_1.pdf
http://www.pmillett.com/Books/crowhurst_basic_2.pdf
http://www.pmillett.com/Books/crowhurst_basic_3.pdf

Pete's online technical resource:
http://www.pmillett.com/technical_books_online.htm

(and I am in no way knocking Nelson Pass, who, besides being a great designer, shares some of his designs gratis with the DIY community via www.passdiy.com and publications like audioXpress).

EDIT^2: Also some interesting reading on amplifiers and hifi stuff in general to be found at:
http://www.sound.westhost.com/articles.htm


I initially thought the question was about (for lack of a better term) transmission line theory... resistance, impedance, capacitative and inductive reactance, the skin effect, etc. -- and why some of the esoteric cable designs make a little bit of sense...

reeltrouble1

ShinAce wrote: »

I still don't think we're talking(or seeing) the same thing.

.

Agreed, It would be nice if Victor would join in, but I doubt it, anyway I will though contact him next week and try to get a better handle on what he was trying to say, so we might discuss his thoughts.

RT1

reeltrouble1

ShinAce wrote: »

Quiescent current is idle current by definition. This is set by the bias of the amplifier. If we take an AB amp, we can bias it into class A operation, or shortchange the bias into class B operation. The topology is the same but the class has changed simply because of bias.

I still don't think we're talking(or seeing) the same thing.

So yes Shin you are correct in thinking of this in terms of bias, good one.

Here is a bit more in depth explanation I received from Victor at BAT.


Hi Ted,

Perhaps the easiest way to think of that (quiscent power) would be analogous to the power amp bias current. As we know, biasing the output stage heavier, as much as possible into Class A, usually improves the sound quality.

The idea is to make any signal-induced current fluctuations very small
compared to the stage bias current, hence reducing its effect on circuit
operation.

One of the notable changes that we made in amplifier design was the
tremendous increase in gain stage power compared to the typical designs.
For instance, a typical gain stage found in many preamps would be running at
2mA or so, while by comparison the VK-50 gain stage ran at 150mA. As result
it is able to drive any cable without any buffer or follower.

The same is true of the power amplifiers, where, for instance, the massive
VK-600 only has TWO gain stages. The traditional designs may have five or
more. This was made possible by - among other things - substantially
increasing the gain stage bias current.

Regards,

Victor

So then by increasing the so called current bias (signal width) you are reducing current fluctuations resulting in better sound.

I received this e-mail from the fellow who bought the BAT VK-500.

He writes:

Ted

It was very nice to meet you and thanks for meeting me 1/2 way. The amp
sounds fantastic, it's the best I have owned. I have had Krells, FPB200c and
FPB300, the KAV250a, I have had McCormack DNA 1 and .5, ATI's Audio Research D400 MkII and this is hands down my favorite. The music comes alive, the micro and macro dynamics are excellent and it is the least fatiguing of all of them. Finally the bass is on par or a bit better than the Krells.


As as you can see his impressions of the sound of this amplifier follows along with Khomenko's explanation of what you can expect to hear from gear designed to deliver a hi-fi experience based upon proper handling of current power both signal and stage. Although the purchaser may not know all the information behind the sound he does know what he hears in relationship to other equipment he has used. I find his description of the micro and macro dynamics with the amplifier particularly pertinant to the topic.

I believe this to be a good example why the simple statements made by manufature's about watts per channel provide little insight into what type of sound you may hear from any given audio component or why confusion about the effect of audio connects, cables, accessories and even the sound of a speaker itself exsists. Furthermore as we have seen on these pages before the ability of amplifier to run in class A with power should be a major consideration in any hi-fi setup.

ShinAce

reeltrouble1 wrote: »

The idea is to make any signal-induced current fluctuations very small
compared to the stage bias current, hence reducing its effect on circuit
operation.

That is a separate idea that warrants further investigation. I'll give it some thought and see where it leads me.

tonyb

I got a brain **** reading all that,not hard to do for me anyway.Has all the excitement of the dynamics of a car tire.Though some is interesting,in the end,just like a car tire,all I want to know is,does it do what it's supposed to do,and will the quality hold up?Nelson Pass is a smart guy,no doubt,his designs have been time tested so it must be safe to say the man knows what he's talking about.Too many variables in sound quality from any one piece of gear,or cables.In the end,we all trust our ears without the needless worry of how or what design made it happen.

reeltrouble1

Tony,

Its not for everyone so I would not worry about it, besides my interest lies with Khomenko's thoughts and designs, the BAT gear is superior to many others I have used. So I am interested in exploring why this is.

Shin, yes it is a seperate thought just analogous to the idle current you mention.

tonyb

Just yankin' your chain a bit Ted,I get it.Bat gear,like other top companies that set the bar above others,they have great designs,use quality parts,have good R&D,and are more concerned about the SQ and the end result of their product.Reputation means something.Here,it's all about sales and keeping costs low.Well,thats a tad off topic,but enjoy the mans thoughts,he may well be right about alot of things.