[Dan Banquer]

A reply to John R:
 Yes John a lot of testing can be automated with an Audio Precision 1. The load however is a different story and that is not included in an Audio Precision One. When my friend set up the tests on his AP1 it took him less than 15 minutes to set up sweeps at about 6 different levels. The sweeps were for THD + N. over power and frequency, and CCIF IM over Power. It took about 5 minutes to perform the tests into an 8 ohm load.
  Similar types of testing are done on amplifiers outside of consumer audio. Makes me wonder, that if it's this easy to do this type of testing why aren't more people doing it?
   We also ran the TIM test and S/N. Total time to do all of this was about 1/2 hour.

[DVV]

A reply to John R:
 Yes John a lot of testing can be automated with an Audio Precision 1. The load however is a different story and that is not included in an Audio Precision One. When my friend set up the tests on his AP1 it took him less than 15 minutes to set up sweeps at about 6 different levels. The sweeps were for THD + N. over power and frequency, and CCIF IM over Power. It took about 5 minutes to perform the tests into an 8 ohm load.
  Similar types of testing are done on amplifiers outside of consumer audio. Makes me wonder, that if it's this easy to do this type of testing why aren't more people doing it?
   We also ran the TIM test and S/N. Total time to do all of this was about 1/2 hour.

Perhaps because such tests would tell a story not to be told?

Cheers,
DVV

[Dan Banquer]

There seem to be two schools of thought here. The first one says it doesn't matter. This is plausible because many loudspeakers aren't linear enough so that you would really be able to appreciate it. The second school of thought says that audiophiles don't want this.
   The fIrst school of thought  ignores the fact that there are some good loudspeakers out there, the second school of thought doesn't want to educate it's readership, and would prefer to keep them ignorant so they can proclaim themselves "The Great Authority"
    Or maybe both schools of thought really go hand in hand?

[nathanm]

I think you electronics wizards here have a unique talent that the rest of us don't have.  You are able to look at a measurement and have an idea in your head as to how that sound will translate through the speakers.  I see it as someone who is able to read sheet music.  I myself cannot read music, so it amazes me when someone can look at a sheet of dots and lines and "play" the tune in their head.  I imagine that you guys who look at scopes and meters have a similar talent of disseminating measurements into sound.  I could look at charts all day and still not have a really solid idea as to how that translates to what I hear.  I imagine it takes a while to learn how what you see on the chart sounds to your ears.

[Dan Banquer]

Nathan: It's not just talent, it's a whole lot of training, experience, studying, reading, and experimenting. What's that old cliche, 99% perspiration, to 1% inspiration. But I can't really say that because I'm bound and detemined to become cliche free.

[hairofthedawg]

I personally think it's mostly experience, which comes with experimentation. Training's important, but I've had a bit of electronics training  and ,although I can make sense of the various instruments used in designing audio equipment, I can't interpret how what I'm looking at will affect how it sounds.

cheers,

Dick

[nathanm]

Okay how about this:

I think you electronics wizards here have unique training, experience, studying, reading, and experimenting that the rest of us don't have.

I guess it depends on one's definiton of talent, if it is something you're born with or something you learn.  Ahh but let us not get mucked up with semantics, this threads' already long enough as is!  

[Ravi]

Sorry to bring this up again, but can Dan, Dejan, or John, (or Hugh or Curt) explain this to me:  Why do tube amps sound so good, despite having more distortion than a highend SS amp in the 15 to 20khz range?  Now I'm speaking about tube amps that are relatively affordable, not 10k tube amps.  Or are there tube amps that have exceptionally low THD in that frequency range and don't require a mortgage to buy?

Is there any viable explanation.  You would think that this distortion would be somewhat audible.  Please refer to my post on the previous page of this thread for more info.

Thanks!

[DVV]

Sorry to bring this up again, but can Dan, Dejan, or John, (or Hugh or Curt) explain this to me:  Why do tube amps sound so good, despite having more distortion than a highend SS amp in the 15 to 20khz range?  Now I'm speaking about tube amps that are relatively affordable, not 10k tube amps.  Or are there tube amps that have exceptionally low THD in that frequency range and don't require a mortgage to buy?

Is there any viable explanation.  You would think that this distortion would be somewhat audible.  Please refer to my post on the previous page of this thread for more info.

Thanks!

In my view, because of three things:

1. They sound good because of the distortion. The thing is that most people will prefer a sound with distorted even harmonics to a sound with no distortion at all. In other words, it's like when a lady tells you you are the world's greatest lover, and you know she's lying to you, but it still feels nice to hear it said;

2. Tubes sound good BY COMPARISON to their opponents, and it is a sad fact that most SS equipment is not designed as it could and should be because somebody's in a rush to deliver the new model, so they seal it with a 2 inch face plate and that's it;

3. If an SS circuit is designed carefully, and is pushed to what it can deliver safely but to what is usually way above what you are normally offered, you will (not might, will) find that it actually beats tubes on all accounts (similarly priced units, that is).

Cheers,
DVV

[JohnR]

Well, a couple of points to add here. One is the theory that low-order even-order distortion actually masks higher-order distortion components. Before you scoff, consider that the ear itself generates several percent of distortion, so there's obviously more going on than just low numbers.

Another is that solid-state amps have tradeoffs too. I owned an AR 100.2 for a while, smooth smooth highs but the bass was as flabby as an obese elephant. That's a 3.5k retail solid state amp.

FWIW

JohnR

[Dan Banquer]

There were some excellent articles in Electronics and Wireless World back in the 70's and 80's. Articles from Moir, and Belcher are some real classics. The latest AP1's are implementing what Belcher was talking about in the 80's from what I understand.

[DVV]

Well, a couple of points to add here. One is the theory that low-order even-order distortion actually masks higher-order distortion components. Before you scoff, consider that the ear itself generates several percent of distortion, so there's obviously more going on than just low numbers.

I'm not sure I understand what's this in aid of? I am not disputing what you say, but I don't see where you are leading to?

Another is that solid-state amps have tradeoffs too. I owned an AR 100.2 for a while, smooth smooth highs but the bass was as flabby as an obese elephant. That's a 3.5k retail solid state amp.

FWIW

JohnR

John, I don't know of any amp, made by anybody, at any time, using any technology, which was not a compromise between this and that at some point. A no compromise amp is another name for a perfect amp, which to the best of my knowledge, does not exist.

I believe this why designing amps is something of an art (well, at least, it can be) - you have to tweak the compromises so that they are all present in about an equal value, so none leads and none lags.

Look at Dan's design - he all but eliminated the electrolytic capacitors from the picture, but compromised the cost of his product by using full electronic regulation. I would venture to guess that this added like say 30-40% to the cost, no small share, but it brought its benefits too.

Anyone who sits down to design an amp, exactly like Dan, has to make his own set of compromises, be it tubes, or transistors, or MOSFETs, analog or digital.

Cheers,
DVV

[JohnR]

Well, a couple of points to add here. One is the theory that low-order even-order distortion actually masks higher-order distortion components. Before you scoff, consider that the ear itself generates several percent of distortion, so there's obviously more going on than just low numbers.

I'm not sure I understand what's this in aid of?

In response to Ravi's question.

It's an intriguing idea, don't you think?

JohnR

[DVV]

Well, a couple of points to add here. One is the theory that low-order even-order distortion actually masks higher-order distortion components. Before you scoff, consider that the ear itself generates several percent of distortion, so there's obviously more going on than just low numbers.

I'm not sure I understand what's this in aid of?

In response to Ravi's question.

It's an intriguing idea, don't you think?

JohnR

In other words, the masking effect at work. I can appreciate that, if it works elsewhere, why not here too?

As for the ear's distortion, I would think it constant no matter what you listen to, where and with what, which levels the playing field.

Cheers,
DVV

[JohnR]

Heh, of course you are right there, I wasn't suggesting that the ear behaves differently according to whether you are listening to a tube or transistor amp  

I notice that certain solid state amps have a small amount of low-order distortion. Dan's amp is 0.1%, AKSA is less than 0.2% across the range. While small, these numbers are still sufficiently large that they look "bad" compared to numbers like 0.001% that you see touted around. I also notice that some manufacturers won't even publish a THD spec of any kind. Perhaps it's because these amps also are carefully designed for maximum sonics and not for the meter, and they are tired of having consumers and reviewers even misinterpreting a number like 0.1% as "100 times worse" than 0.001%.

So, taking the leap into dangerous waters, I wonder whether in fact the presence of this amount of H2 is beneficial to the subjective quality of an amplifier.

JohnR

[Dan Banquer]

I think neither. At present the audibility of second order distortion is about 1% to 3%. From my experience it does not mask higher order harmonic distortion.

[Dan Banquer]

I am still saddened to see your comments about meters and perception of sound quality. My perception of this is that there is important data to be taken. It's not, however,talked about really at all in consumer electronics these days. Maybe that will change in a few years. I hope as you continue with your DIY that you will start to get curious and go deeper into the subject at hand.

[JohnR]

I am still saddened to see your comments about meters and perception of sound quality. My perception of this is that there is important data to be taken. It's not, however,talked about really at all in consumer electronics these days. Maybe that will change in a few years. I hope as you continue with your DIY that you will start to get curious and go deeper into the subject at hand.

Be as sad as you like, Dan, but do it for the right reasons! What on earth did I say wrong? Here I am trying to talk about distortion spectra and what measurements are meaningful, and you're putting me down for it!

I think you're probably just misinterpreting what I'm saying. Nothing new there

JohnR

[JohnR]

I think neither. At present the audibility of second order distortion is about 1% to 3%. From my experience it does not mask higher order harmonic distortion.

I'm not sure what the "neither" is. But out of curisiosity, under what conditions was the audibility of H2 determined?

JohnR

[DVV]

Heh, of course you are right there, I wasn't suggesting that the ear behaves differently according to whether you are listening to a tube or transistor amp  

I notice that certain solid state amps have a small amount of low-order distortion. Dan's amp is 0.1%, AKSA is less than 0.2% across the range. While small, these numbers are still sufficiently large that they look "bad" compared to numbers like 0.001% that you see touted around. I also notice that some manufacturers won't even publish a THD spec of any kind. Perhaps it's because these amps also are carefully designed for maximum sonics and not for the meter, and they are tired of having consumers and reviewers even misinterpreting a number like 0.1% as "100 times worse" than 0.001%.

As I have pointed out several times, it's not at all the same thing HOW you get such low, low numbers. The trouble is that more oftne than not, they are achieved by raping the amp with too much negative feedback.

So, taking the leap into dangerous waters, I wonder whether in fact the presence of this amount of H2 is beneficial to the subjective quality of an amplifier.

JohnR

In my truly honest opinion, John, I think not at all. To my mind, 0.1% across the range is the realistic limit of what a good designer should aim for and call it a day. I would call this a realistically achievable number, possible to obtain with careful and balanced design, and with token negative feedback, where NFB is actually used to make a good think a little better (i.e. the amp does not require NFB for stability).

Obviously, I completely agree with Dan and Hugh on this point.

Without misreading your message, or taking it out of context, I would once again stress that the best of designs have all forms of distortion contributing EQUALLY, and are therefore surpressed equally by negative feedback. I think few people are aware of this simple truth, which has been with us for over 30 years now.

Most designers aim for reducing those forms of distortion they know magazines will measure, which all too often leads to overlooking or underrating other forms of distortion present in their designs.

An Italian magazine "Suono" (Sound) had a wonderful test they called TriTIM; it would show in 3D, much like the waterfall diagrams given for speakers, how an amplifier behaves in the -3 ... 0 ... +3 dB power range (i.e. half-rated-+50% power range) when excited by 15/16 kHz test tones. Oh boy, did that test catch many a model with its pants down! You want to see how some famous names fared - terrible.

Cheers,
DVV