If anyone knows, I'm curious:
>>Power Paradigm is the method of design, test and measurement<<
>>Crossover design rules<<
When did the term "Power Paradigm" first enter the lexicon by name? Where are the methods for test and measurement codified? Where are the crossover design rules described?
The use of the word 'Paradigm', as described in the opening statements of my article, is because a paradigm is used to describe a platform of thought. Anything outside of that platform is considered blasphemy. Look at some of the responses here and elsewhere and you see what I mean- the comments are derisive, using a proof from the Voltage Paradigm. As long as the participants stay in their box, they can't see what is outside of it.
When were the characteristics of power paradigm amps first described as a preferred way to drive loudspeakers compared with voltage paradigm amps?
Essentially in my article, as prior to that the industry was stumbling around with the 'equipment matching' conversation- and a lot of audiophiles were flushing perfectly good money down the loo.
What output impedance range defines a power paradigm amp? >>1-20 ohms is typical<< What is the ideal impedance? Is it flat over the audible frequency range? If not, is there an ideal characteristic?
The output impedance should be flat over the audio range. Most amps I see (SETs, low power triode push-pull zero feedback amps, our OTLs, low or zero feedback PP tube amps, single-ended transistor amps) have an output impedance greater than one ohm and less than about 10. Damping factor is obviously low, and so the amp should be paired with a speaker that does not require high damping (FWIW, no speaker needs more than 20:1, see 'Missing Link in Speaker Operation' by D.J.Tomcik (one of the EV engineers).
Is 2nd-order distortion (e.g. 10%) merely inconsequential, or is it a positive attribute in the rendering of sound? If it is positive, is it desirable to have more than 10%? Is there an optimal value? How much is too much? What about other even-order harmonics? What is the ideal relationship of these to the 2nd harmonic? Falling? Falling fast? None?
None is preferred if possible without adding to odd orders.
If a 30w Power Paradigm amp makes 5% 2nd harmonic distortion at 10w, what would it ideally be at 20w? Same? More? Less? Doesn't matter? Is there an optimum slope to the curve or relationship between 2nd harmonic and power output defined in the Power Paradigm school?
The best distortion characteristic seems to be one wherein the distortion linearly decreases to unmeasurable as power is decreased from full output.
Excellent questions. Another: Are voltage paradigm and power paradigm clearly defined engineering terms that are broadly accepted and applied, or are they more proprietary language? I ask because a Google search turns up little, and leads straight to Atmasphere. Perhaps these are terms of your own creation?
Tim
As mentioned in the opening statements of my article, I used the word 'Paradigm' for a reason that had nothing to do with engineering.
Really? Exactly what did I miss and what incorrect assumptions did I make? You basically rehashed everything you already said.
Yes. This is because you seemed to either not read the ariticle very carefully, or have poor reading comprehension. I outlined some of the faulty assumptions in my first post.
I have to be honest and tell you that I’m very tired of you accusing me of “logical fallacies” and “strawman arguments.” I’ve already pointed out several times how you changed my words in a previous statement I made so you could turn around and accuse me of a strawman argument when there was none until you created a reverse strawman argument. And if you haven’t figured out by now that I did read and understand your paper and that it would have probably been a better choice not to be so condescending in your tone and remarks that we might actually have a more desirable conversation.
Fine. I will take you at your word.
Ralph-Please tell us if there are any tube amplifiers with output transformers on the market with an output impedance of 5 ohms or greater. If you want to call an amplifier with an output impedance of 10-20 ohms moderate, do you mean moderate for an OTL? And even though I know you said 1-20 ohms would be considered moderate, I picked the higher end of your band and just left 10-20 ohms because that wouldn’t be considered a moderate output for a tube amp with output transformers nor is it common to all OTL amps either. Both the Graaf GM200 and the Transcendent Audio T8 OTL amps to list two examples have an output impedance of under 1 ohm.
Leave OTLs out of this and maybe it will be easier to understand. Although the ones we make do have a fairly high output impedance, that is not true of all OTLs by any means. What we are talking about is an amplifier with a high output impedance because it has no feedback. That can be any sort of technology that qualifies and could be solid state.
What speaker company advertises their speakers as “power paradigm” speakers and who advertises their speakers as “voltage paradigm” speakers? What high-end magazine or audio journal classifies speakers in this manner? Are there any? From time to time you will see measurements of a speaker by JA in SP and he will warn people the speaker under test will best be driven by a SS amp due to the demands it will place on a power amplifier. I think if speakers were strictly designed and built to be “power paradigm” speakers or “voltage paradigm” speakers, they would be advertised and sold that way and we would see special categories in TAS, Stereophile, and other magazines in their recommended component special issues for speakers in these categories.
None. What my article is about is raising consiousness, and at some point perhaps a speaker manufacturer might advertise that their product requires an amp capable of constant power.
SP ranks their speakers according to Class A, Class A limited low frequency, Class B, and Class C. Unless I missed it, there is no breakout for “power paradigm” speakers and “voltage paradigm” speakers. The only information/recommendation I see coming from the majority of speaker companies regarding amplifier choice are the standard efficiency, nominal impedance, minimum recommended power, and maximum recommended power.
If speaker companies truly designed and built their speakers to only be used by “power paradigm” amplifiers or “voltage paradigm” amplifiers, don’t you think they would tell us that? I think the real truth is that most speakers can be happily driven by either SS or tube amps with output transformers, but the number of speakers that can be driven by an OTL amp with an output impedance of 10 ohms or greater with high fidelity will be a much smaller subset of speakers that can be driven very well (meaning with high fidelity) by either a SS amp or a tube amp with output transformers.
Of course they don't, but if you indeed read the article- what did I say?? 'Ask the designer- it is a matter of intention', to paraphrase. What is the reference amp the designer uses? mep, you clearly have a bias towards transistors, would you buy a speaker wherein the designer was clear that he used tube amps exclusively? Might your refusal to buy that speaker be because it might not sound right with your amp??
First of all, I don’t know how a speaker could be “designed to expect a certain behavior out of the amp if the impedance is lower.” Unless the speaker company specifically recommends an amplifier for use with their speaker, how would they possibly know what to expect from all of the possible amplifier choices? With regards to the Wilson example you have used repeatedly, Wilson Audio does not market their speakers as power paradigm speakers that should be driven by tube amplifiers. Wilson provides some adjustments you can make to their speakers in order to accommodate SS or tube amplifiers.
That 2KHz trap worked the way Wilson designed it. He clearly had an amp with a higher output impedance when he did so ('intention' that I mentioned earlier). If you put a transistor amp on a 2 ohm load, it will **try** (may not be successful) to put out 4X greater power than it would into 8 ohms. This is why such amps sounded bright on that speaker, which I already explained. What Wilson could predict is that the speaker would sound fine with a tube amp. It was in practice also quite predictable, and that is one of the major points of this discussion.
Do you have any papers you could reference with regards to McIntosh being leaders in the voltage paradigm? Since both McIntosh tube amps and all but their cheapest SS amps use output transformers, that puts them in the “power paradigm” camp and not the “voltage paradigm” camp. And it would only follow that if McIntosh amps were designed following the rules of the “power paradigm” that their speakers would be designed the same way.
You are mistaken- output transformers have nothing to do with this one way or the other. It is more about global loop feedback.
I think it would be a more accurate statement to say that quite often SE triode tube amplifiers operate with little or no feedback. It’s far less common to see push-pull tube amps use zero feedback than it is to see SE amps using zero feedback.
This is true, but I have such an animal (P-P class A amp using type 45s) at home. Just because they are rare does not disqualify them. 20 years ago SETs were rare and by most counts still are, allthough they are a lot more popular now than 20 years ago.
He commented on it all right. Apparently right back to Ralph and he didn't agree with him. Here is the post you linked to and it is very interesting to say the least:
Posted by Bob Cordell (A) on September 12, 2006 at 13:27:16
Ralph,
Thanks for your interest in the workshops we will be holding. I think a couple of them will be relevant to your question. First, we will be doing measurements on amplifiers and in so doing will discuss and speculate on how measurements do or do not correlate to the sound. In the amplifier listening workshop, we will compare the sound of a vacuum tube amplifier with that of a solid state amplifier; because the output impedance of the vacuum tube amplifier is a bit higher, one could argue that it falls closer to what you describe as the constant power paradigm. Finally, we will provide a demonstration of the average and peak power (referred to 8 ohms) levels on well-recorded music playing into a given loudspeaker. Our plan is to display rms average and peak value of average power simultaneously on digital readouts calibrated in watts average power into 8 ohms. This is intended to illustrate the crest factor of the program, and by extension the likelihood of whether one's amplifier is clipping.
Now to the specifics of your question. I have to say that I don't subscribe to so-called constant voltage and constant-power paradigms as you described them. The constant power paradigm is particularly problematic, since with most loudspeakers, if you deliver a truly constant power to the device, its frequency response will be very far from flat. Indeed, if you look at the impedance curve, you will see for many speakers a large rise, sometimes to tens of ohms, at the bass resonant frequency and again at the crossover frequency. Such impedance rises suggest that the true power efficiecy of the loudspeaker is significantly increased at these frequencies. Feeding constant power to the loudspeaker at these frequencies would result in both a voltage rise at the terminals and an SPL rise at the acoustic output. The bottom line is that most speakers are designed for, and voiced for, a constant voltage input (regardless of the power that that happens to correspond to).
I know of no audio power amplifier that produces constant power into a widely varying frequency-dependent speaker input impedance. You may get a little closer to that with an amplifier with a higher output impedance, but still not a lot closer. Closer enough to alter the sound, however. Even the constant current amplifiers that some have proposed are not constant power.
I also don't see stereotyping so-called subjectivistis and objectivists into constant power and constant voltage camps, respectively. This would be similar to categorizing bottleheads as all subjectivists and solid state afficiandos as all objectivists - something that seems wrong to me.
I hope I have addressed your question, and hope you can join us at RMAF for our workshops.
B
mep, your Appeal to Authority here is a classic Logical Fallacy. Just because Bob didn't know of the amp in bold does not mean it does not exist. For that matter I don't know of one either, in strict terms, but if we speak in terms of decibels than I can name quite a few. Further, the constant current amps he mentions of course are not constant power. The comment actually points to his ignorance on the topic. His closing argument is also a Strawman- I never made the stereotype he suggests.
If one "Power Paradigm" amp's source impedance is 20 ohms and another is 1 ohm, what can you as a designer rely on for consistency? OTOH, all Voltage Paradigm amps have very low output impedances, thus removing them from the equation as a variable in how the speaker will react, i.e. sound. If the Wilson Watt/Puppy needs a 2 kHz trap to tailor the sound, that is easily immunized from the low source impedance of a typical power amp. I'd be very surprised if Wilson failed to design it accordingly, because to rely on the soft and uncertain source impedance of a high impedance amp means the sonic result would be left to serendipity. That does not sound like Dave Wilson to me (a.k.a detail control freak).
Having a wide range of load tolerance is not an idiosyncrasy in my book. Eliminating unpredictable and uncontrollable variables from a system is a good thing.
Here is the problem you are up against. It has to do with the human hearing perceptual rules. We already know that the brain has certain tipping points and also that the brain interprets distortion as tonality (which is why a trace amount of odd ordered harmonic distortion in an amplifier can result in brightness). The brain will favor distortion as tonality over actual frequency response if the distortion is high enough. This occurs at different points depending on the harmonic itself and other factors. IOW, the Voltage Paradigm insures flat frequency response, but it may not actually sound that way due to distortion issues.
Exactly. One could summarize Cordell's argument as follows:
- If an amplifier actually did deliver constant power to a speaker regardless of impedance, the system frequency response would be severely affected.
- No such amplifiers exist anyway.
- An amplifier having a high output impedance is not even a good approximation to one that delivers constant power to a load of varying impedance.
And Cordell is incorrect on all points, if we are to take the above as what he is saying. The MA-1 for example is 140 watts into 8 ohms, 150 into 16 and 144 into 30 ohms and 135 into 4 ohms. In terms of decibels that's a pretty flat power curve. The MA-2 is 220 watts into 4, 8 and 218 watts into 16 ohms. That's even flatter. So points 1 and 2 are moot, its really point 3 that might be an issues, except that one is a Red Herring, see
http://www.nizkor.org/features/fallacies/red-herring.html , IOW it does not matter one way or the other.
