Da high slew rate heresy is almost all due to da false prophet Matti Otala. I won't go into why his stuff is liquid BS. Bob Cordell does a good job in his books, articles & several threads here.
What I have done is DBLTs on his designs, some of which were commercialised under his grubby hands.
Guess what? They do TERRIBLY on DBLTs even to da man in da street.
A clear preference, even by a true Golden Pinnae in a single series of DBLTs is subjective.
If everyone in a series of DBLTs all prefer the same thing, this is objective data, ie a fact .. tempered by statistical analysis & other sparse data BS 🙂
Each of da IMD tests relate to and measure a particular order of non-linearity which is also measured by a 'single harmonic' of a 'THD' measurement ... provided the DUT and measuring system has sufficient bandwidth bla bla .. and other caveats. I used to have all this worked out with da scaling factors and jotted in our manual of the B&K 1902 distortion analyser which could do all da tests.
Excuse me if my single remaining brain cell ken no longer du da maths. If someone with more intact brain cells wants to do it, I'm sure we'd all appreciate this 🙂
How do you do this? Are you using commercial recordings? Which ones?
Are all your "large and diverse panel of listeners" recording engineers? I had a couple in my A DBLT panel including Bat Ears who would insist on using his own recordings. I'm guilty of this prima donna behaviour myself 🙂
Actually one very common tube topology is VERY prone to yucky 'sticking' behaviour on overload ... though it isn't 'rail sticking'. Da Mullard 5-20 variants which includes practically all the last of da great tube amps including Carver's Silver 7, don't do this.
Valve amps CAN sound better than transistors but it's important to understand why. Bob Carver made his first 100W SS amp and thought his old 50W valve amp sounded better. What he found was that a valve amp can deliver a LOT more volts into high Z. This led him to the Phase Linear 700; at that time the biggest amp you can buy.
I had a similar experience when the $$$ Luxman (?) valve amp was all the rage in the UK. I dug up an old LEAK 50W amp which did indeed sound better than the Luxman and the SS stuff we used at the time. The end result of this was 1000W/channel for our listening room and 200W/channel for my own use.
IMHO, Bob Carver was the most innovative amp designer ever and pioneered the most stuff that led to better sound. See da Stereophile Challenge mentioned earlier. Pity his stuff tended to release da Holy Smoke a bit too often 🙁
Bob Carver is still alive and still could pop out some excellent amplifier idea at any time.
He might not appreciate the was status.
He might not appreciate the was status.
If the MJL4281 had been around when the PL700 was introduced, people would have been singing its praises concerning reliability. Would have been nice to be able to get them in the metal can, too.
To put them in old PL700’s (and amplifiers like them) so they don’t go szhekx in the night again.
I'm not interested in THD measurement because for me it is relative simple to achieve low distortion amplifier in all audio frequency. I want to know how to define measurement related to dynamic condition. I found some amplifiers sound good when play with few music instruments and slow beat, but if the source have many music instruments, fast beat, and high dynamic then they sound bad or not good enough. Some sources expose the weakness of the amplifiers. I also found some amplifiers after play music in around one hour, then the stereo image going to bad. In my country, people really nice and welcome, so we can listen their audio setup, even we can bring our amplifier or speaker to their home and listen the music.
No, it think it would have to be specific recordings with a reference system to record and playback... and reference recording and playback rooms. Ideally, we'd like to make it so only the differences can be measured. Take the live performance, record it and then record the reproduction and see the differences. Have listeners at both live and recorded playback instances.
For the listeners, it would have to be a wide group. Not necessarily Golden Ears nor "experts", just a wide enough group that would filter the individual preferences. Hopefully the listeners have some experience with the music being used.
Things like time frames will be important... perhaps do the experiment within 30 minutes. 10 minutes to play/record the original performance with listeners in place... then 10 minutes playing it back while a recording of the playback is made. Then gather the "opinions" of the listening panel and compare the original recording with the 2nd recording.
Perhaps also make the playback/recording equipment adjustable... location, distortion components, etc... the way the First Watt SIT-1 and the Mini ACA amps allow you to select different types of distortion components.
Note that this is my first take about this, I'm sure there are many more steps, more iterations to make this work... but the bottom line is that we are trying to correlate WHY some components sound better than others and come out with measurements and psycho-acoustic theory to explain it.
Ultimately we want to know which aspects of sound reproduction are important to recreating the best facsimile of the actual event and develop some means to create the concomitant measurement techniques.
That's what I think "quality" of an amp is all about.
Last edited:
This is simply not practically possible - forget it. I mean in context of what is basically discussed here.
//
There are two types of activities - to design, to materialize design results, or to come up with criteria for evaluating design results.
In the first case (design and materialization of results), the system of criteria arises as a by-product and over time acquires a certain integrity and meaning.
In the second case, the process is guaranteed to be endless because it is based not on practice but, relatively speaking, on desires.
The DIY method of design and materialization is very interesting because it excludes the commercial component, the designer is free to concentrate on what he considers interesting, beautiful and suitable for solving his problems, without the strong restrictions of mass production, sales and service.
For example, PMA's point of view is that good dynamic characteristics of an amplifier when operating a composite (not purely resistive) load, combined with simplicity and even elegance of circuit design, which ensures design reliability for many years, it is more important than sub-PPM THD. This is his path as a designer, and along this path he came to a truly beautiful and interesting design.
There are designers here who are in love with amplifiers without global feedback; they, for example, hone the output diamond with enviable persistence.
There are designers who create very beautiful asymmetrical SE amplifiers.
There are designers who specialize in circlotrons.
Each of these designers has its own system of evaluation and priorities of individual criteria.
And this is precisely what is wonderful, it saves us from the dull, win-win commercial monotony (at a short stage) of “take a ready-made class D module with ppm distortion and concentrate on the commercial component”.
No one needs or is interested in synthetic criteria-based assessment systems precisely because they are synthetic artificial constructions behind which there is usually no practice.
Commercial amplifier designers have their own criteria systems, each company has its own, no one will ever show them to you, but these systems have one thing in common - they work. You can show off your contempt for Bose, for example (this probably says something, but I don’t know exactly what), but I’m already on my fourth car with Bose acoustics, and I can’t say that I’m dissatisfied. For me, this means that Bose designers know and can do it.
DIY designers have their own criteria systems. And there are as many of these systems as there are DIY designers. And that's just wonderful.
Dat klopt!
In the first case (design and materialization of results), the system of criteria arises as a by-product and over time acquires a certain integrity and meaning.
In the second case, the process is guaranteed to be endless because it is based not on practice but, relatively speaking, on desires.
The DIY method of design and materialization is very interesting because it excludes the commercial component, the designer is free to concentrate on what he considers interesting, beautiful and suitable for solving his problems, without the strong restrictions of mass production, sales and service.
For example, PMA's point of view is that good dynamic characteristics of an amplifier when operating a composite (not purely resistive) load, combined with simplicity and even elegance of circuit design, which ensures design reliability for many years, it is more important than sub-PPM THD. This is his path as a designer, and along this path he came to a truly beautiful and interesting design.
There are designers here who are in love with amplifiers without global feedback; they, for example, hone the output diamond with enviable persistence.
There are designers who create very beautiful asymmetrical SE amplifiers.
There are designers who specialize in circlotrons.
Each of these designers has its own system of evaluation and priorities of individual criteria.
And this is precisely what is wonderful, it saves us from the dull, win-win commercial monotony (at a short stage) of “take a ready-made class D module with ppm distortion and concentrate on the commercial component”.
No one needs or is interested in synthetic criteria-based assessment systems precisely because they are synthetic artificial constructions behind which there is usually no practice.
Commercial amplifier designers have their own criteria systems, each company has its own, no one will ever show them to you, but these systems have one thing in common - they work. You can show off your contempt for Bose, for example (this probably says something, but I don’t know exactly what), but I’m already on my fourth car with Bose acoustics, and I can’t say that I’m dissatisfied. For me, this means that Bose designers know and can do it.
DIY designers have their own criteria systems. And there are as many of these systems as there are DIY designers. And that's just wonderful.
Dat klopt!
Last edited:
Everything you say is true. However, gross design failures in any one of the amplifier types (and you can apply this to any other non-audio activity you care to mention where there are multiple ways to achieve a similar end objective) are a separate issue and this is what is being addressed. For this reason, the proposed assessment does not go anywhere near subjective sound ratings but looks at the technical performance across a range of parameters, rather than a focus on THD, which will affect sound quality negatively, however you wish to define that. The assessment rates different amplifier technologies separately to avoid lumbering anything without high feedback as below standard. That cannot be the case when there are very credible zero global feedback (see Robert Cordell 'Designing Audio Power Amplifiers) and tube amplifiers with good specs within their technological capabilities. At this point, the technical parameters are
Hum and noise only (I have not met one amplifier builder/manufacturer where mains hum was a saleable feature, no matter what the amplifier type)
Complex load drive capability problems (same comment as above - as far as I know, this type of behaviour has never been a sales feature)
THD+N (who likes hum and excessive THD?)*
IMD at full power (who wants an excessive spray of low and high-order harmonics on music crescendos?)
Capacitive load drive capability (oscillation has never been a benefit)
Clipping behaviour (who wants audible overhang and/or oscillation when an amp exits clipping?)
If you like hum, noise, sticky rail, oscillation, or deleterious response anomalies when driving complex loads then the proposed assessment is not for you. If you think you can raise your game by attending to most or all of the areas listed above where problems can arise, then this should help you.
*There is a case to be made about THD harmonic content and how that is measured, but there is no consensus on how that will done at this stage.
Hum and noise only (I have not met one amplifier builder/manufacturer where mains hum was a saleable feature, no matter what the amplifier type)
Complex load drive capability problems (same comment as above - as far as I know, this type of behaviour has never been a sales feature)
THD+N (who likes hum and excessive THD?)*
IMD at full power (who wants an excessive spray of low and high-order harmonics on music crescendos?)
Capacitive load drive capability (oscillation has never been a benefit)
Clipping behaviour (who wants audible overhang and/or oscillation when an amp exits clipping?)
If you like hum, noise, sticky rail, oscillation, or deleterious response anomalies when driving complex loads then the proposed assessment is not for you. If you think you can raise your game by attending to most or all of the areas listed above where problems can arise, then this should help you.
*There is a case to be made about THD harmonic content and how that is measured, but there is no consensus on how that will done at this stage.
Beautifully stated. And that is exactly why I've repeatedly asked for whom this "rating / assessment / criteria" is intended. The originator claims that "almost every" designer is in a race for sub-ppm distortions "at the expense" of other criteria that are also important. They seem to be concerned with the commercial market and comparing this "new" metric to things like the SINAD published by ASR.
They've provided ONE piece of supporting evidence for the claim of sub-ppm distortions at the expense of "reliability". Perhaps there are other examples, but more importantly, you've eloquently described the realities of both DIY and commercial motivation in their most general sense.
A circuit designer for a commercial product is going to do what their boss / marketing research tells them to do, unless they're a small / boutique shop. Likewise the small / boutique shop designers likely don't care one iota re: some new "rating". They have their own (equally valid) sets of merits for their designs which might even be associated with their very name. They're not going to (likely) change their design criteria. They'll change their reliability criteria based on knowledge and/or returns/complaints. They'll change their performance criteria around pre-set targets and component availability along with their desired design and cost criteria.
From a DIY standpoint, as I've repeatedly asked... assign whatever scoring is the fad of the day, but ALSO provide the raw data. A standardized test suite (to me) has the most value. Some algorithm regarding the perception of the overall "goodness" of a design and totality of "quality" has very little value to me, and I doubt much value to the "market". I hope to be proven categorically wrong.
To me, the beauty is in the ability to take the data from standardized testing and craft it into those myriad of 'systems' based upon our own needs and wants.
Again... I support the effort fully b/c it supports and will document (I hope) a standardized test suite approachable to DIYers.
Last edited:
Reliability/MTBF is difficult to measure till you're already entered the market, and it can change over time due to production changes.
Also low distortion does not have to come at the expense of other performance metrics, it just takes sensible design, ~10ppm isn't hard to attain for instance with straightforward architecture into 8 ohms for class B. Same for low noise, this isn't rocket science.
One problem throughout this thread is that the word "quality" is almost meaningless, or rather has many different meanings from ergonomics to MTBF to audio performance to power efficiency to flexiblity of configuration to value-for-money to "looks stylish", etc...
Also low distortion does not have to come at the expense of other performance metrics, it just takes sensible design, ~10ppm isn't hard to attain for instance with straightforward architecture into 8 ohms for class B. Same for low noise, this isn't rocket science.
One problem throughout this thread is that the word "quality" is almost meaningless, or rather has many different meanings from ergonomics to MTBF to audio performance to power efficiency to flexiblity of configuration to value-for-money to "looks stylish", etc...