Class A hybrid power amp up and running

I see Serge posted as I typed. Good point on the telephony yes. Only where lines are miles long can they be classic transmission lines with a characteristic impedance at audio frequencies.

 

I was not associating return loss with damping factor, the two were just conflated and thus the mis - understanding .

Classically, with electro - mechanical telephone exchanges, slow to release relays were created by placing a set of contacts across a relay, as it was de-energised. This short enabled the back EMF - the field of relay coil to extend the hold time, while it decayed - not trying to teach granny to suck eggs. The same principle works on loudspeaker cone movement when the signal - audio - changes. Obviously, the closer an amplifiers output impedance is to zero, the greater the arresting force, on the cone. So to suggest a unity damping factor 1 to 1 implies the output impedance of the amplifier is equal to that of the loudspeaker. That is nonsense in my opinion and in the case of valve amplifiers, the output transformer forms part of the overall load, along with the loudspeaker. So a valve stage will still have poor damping effect on cone inertia.

However, with regard to Return - loss, classically 15db is a nominally accepted figure within the communications world. At the start of the mobile telephone revolution, that was seen as the gold - standard. But that did not permit systems to function with much more than 4 channels per site. When I finished my research and implementation on live networks, it was up to 24 channels. What I discovered was, this was no where near a good enough match, to keep products low enough - products that were not a function of even the channel concerned.

Throughout the industry, matching was paramount between units and the better the match the better the performance. I speculate that the real difference between the performance of cutting - edge amplifiers and the bulk of the rest, is the care taken in matching components and sections, such that there are fewer spurious products which poach performance due to less than adequate optimization. Any device, not working into a optimized impedance, will have a poor return loss, which can depend upon the frequency at which it appears, causing phase delay and cancelations around the region where cross - over to other drivers occur. It is very easy to tune a filter for a flat pass band. But when one looks at it's return loss things may not look quite so rosy. At the same time, it is good idea to look at what happens to "Out of band products" in such circumstances. It is very easy to produce an amplifier which is unconditionally stable, into 8 ohms - there is no angle. It is a wholly different kettle of fish when one tries to do the same with a complex impedance.

 

Indeed this is where we disagree fundamentally! I consider the differences between well measuring amplifiers to sometimes be as large as those between speakers.... and often of greater importance because the ear tends to be more forgiving of simple colourations etc than to the often "nasty" sounding problems that amplifiers can introduce. I am aware of the Quad and Colloms tests yes but consider them spurious as the differences subjectively between a pair of Quad II's and a 405 are chalk and cheese. I was not aware of the Behringer A500 against Levinson tests but as I've said before I consider the A500 one of the worst sounding amplifiers I've ever heard so can only conclude the tester was half deaf! I would take Kef Codas driven by the Levinson over Quad ESL63's driven by the A500 every time as a very revealing speaker would reveal the true awfulness of the A500.

There seems little point in protracted debate on this matter as we have both made our opinions clear in the past... you think it as crazy to think amplifiers sound different as to think fuses sound different (at least we can agree on the fuses!)... and I think the differences between the A500 and Levinson likely to be of similar magnitude to those between said Coda's and ESL63's and so huge one would have to be deaf to not find it blindingly obvious.

I respect your technical knowledge and enjoy (most!) debate with you but on the more subjective matters (ultimately all that matters is the illusion of well reproduced music and music itself must by very nature be subjective as is all art) I'm aghast at your opinions I'm afraid.

I find it intriguing that someone who believes all hi fi sounds the same (other than speakers? Or do all speakers flat within +/- 3dB over 40Hz to 18KHz sound the same as well?) has such an interest in the subject and made a career out of it!

It strikes me that were you "in charge of hi fi" we would have never progressed beyond say the Leak Delta 70 or NAD 3020 as both measure <0.1% THD and adequately cover 20 Hz - 20KHz. You will no doubt consider all the R+D I've done into phono stages a waste of time as the ubiquitous "single op amp application note special" used in most budget gear can measure very well

The human ear/brain interface and processing are not well understood by science and I wonder if maybe some people who would pass an NHS hearing test are nevertheless unable to hear things like "muddiness", "harshness", stereo imaging and depth perspective etc etc in a similar way that some people are colour blind??.... It's a possibility anyway

There's an irony to all this of course in that those who reckon there is a huge difference between the sound of different wire and even fuses in mains plugs will no doubt regard my opinions on this in much the same way as I regard the opinions of the "all amplifiers sound the same" brigade! The fact that I cannot say why similar measuring amps sound different yet insist they do sound different will be grist to the mill I'm sure FWIW on this one an amplifier has maybe 100+ components per channel, some parts have gain, which is not linear, non are literally perfect, negative feedback means some part of the signal goes around the loop more than once etc etc... ie it's an almost infinitely more complicated physical mechanism/process than a metre of wire which at audio frequencies can be regarded as perfect within Ohm's Law

 

I think a touch of humility would not be mis - placed, or is it your purpose to find fault in everyone's view, which runs counter to yours. Your response to a perfectly reasonable observation is frankly offensive. Whilst you crow about your undoubted achievements, you provide no evidence to support your claims. If you want to debate Hifi, then fine. But, if you simply want to have a "Pop" at everyone, then I'm out. Who, or what was it that created that huge chip on your shoulder?

 

Says a man who insists that non existent "return loss" is the mystery factor that no one else understands....

Debate by its very nature means some will be in favour of and some against the matter being debated. I am not someone who will simply shut up and go away every time someone disagrees with me! Instead I argue my point and try to show why I believe my POV to be correct.

Whilst I'm certainly not unique in my degree of experience in this field I'd wager that there can be few who have designed, built, measured and listened to as many amplifiers over so many decades as I have and this experience has taught me that it is FACT that well measuring amplifiers can sound different to one another.

I've known Serge (through hi fi forums only) for years and I know he is used to people robustly disagreeing with him on this very matter as on hi fi forums in general virtually no one agrees with him on this! ASR is likely the only exception.... I doubt he is offended by my strongly refuting his view that amps sound the same (I hope not anyway) and my "leg pulling" but to me the idea that all well measuring amps sound the same is as wrong as mains fuses changing the sound

 

I would be convinced that similarly measuring amplifiers can sound different if this was the result of properly conducted double blind listening tests, not sighted evaluations. So far, all Jez has said is that HE has listened to many amplifiers and he's satisfied that differences exist.I'm not convinced, and those admittedly few blind listening tests published indicate no such differences.

My own experience of blind listening tests also show no differences, but in my case, it could be expectation bias. I don't expect to hear a difference, so I don't. Those that claim to hear a difference need to do so blind, level matched and with a sufficient number of trials for statistical validity. Unfortunately, it's hard to do this with sufficient rigour, which is why it's very rarely done, and anyway, it's in no manufacturer's commercial interest to prove that their amps are no different to any other.

S

 

Personally I consider the whole blind/double blind thing pointless in the context of evaluating equipment you own and have no "skin in the game"/commercial interest over the results. The kind of differences I'm talking about are obvious without going to these extremes.

Do you have a range of amplifiers available and a decent (don't have to be SOTA) pair of passive speakers in your collection?

 

No surprise there then - is everyone wrong then? I reiterate, match is important and if indeed there are differences, it is most likely due to an amplifiers output stages and how they react to complex impedances - Loudspeakers and Cross - over networks. But I for the life of me, can't think of any of the major makers who aren't aware of this and their designs should perform with a wide range of loudspeakers currently on the market. Plus, one cannot accuse a whole industry to universal dishonesty. True, there are times when things that should be said, are left unsaid. But it is up to the punter to to do his own research and those that spend money on top - end Hifi are not all fools.

How about stacking one of your designs up against a comparable commercial equivalent? I am sure it could and would be found wanting, if folks use the same criteria to judge it as you do to others.

Serge is right when he says everything you have said; to date, is anecdotal. It will do you no harm to provide firm evidence to support your opinions - which is all they are. However good you may be as an Audio Engineer, it is being destroyed by your pig -headedness when it comes to providing proof of your assertions.

 

It's very easy to say that differences between amplifiers are obvious so blind testing is unnecessary, and yet, all such 'obvious' differences disappear when blind testing. Why is that?

Is it that blind testing, the Gold Standard in so many other fields, doesn't work for audio? Or is it more likely that the differences, so obvious when heard sighted, just aren't there?

S

 

There are non so deaf as they who will not hear... and in Hodges case don't have a clue what they are talking about anyway.

No point in further discussion on the matter if you two don't recognise subjective judgement is ultimately all that matters... maybe you should try "measuring" some Mozart to "prove" it's "better" than Beethoven

Everything that matters about this amplifier as far as further improvements and developments go will be of a subjective nature so I'm out of this thread.

 

"No point in further discussion on the matter if you two don't recognise subjective judgement is ultimately all that matters... maybe you should try "measuring" some Mozart to "prove" it's "better" than Beethoven"

No, subjective opinions may matter TO YOU. They don't matter to anyone else, so there's no point in expressing them more widely.

That's why whether Mozart or Beethoven was the better composer is also equally daft. One may be better TO YOU, than to me, so who cares, apart from you?

That's why I have zero interest in subjective reviews, whether of films, books, restaurants or indeed HiFi. If something can't be meaningfully measured, like whether Mozart is better than Beethoven, or Bordeaux is better than Burgundy, then I buy what I like. However, with anything that can be meaningfully measured, like HiFi, I buy on the numbers, as those are the same for everyone.

S.

 

"Steady Boyo", We mustn't go too far! Burgundy is most definitely better than Bordeaux. Not when one talks about generic Burgundy or Bordeaux, as there is little to choose between them.

However, If you have ever drunk, not tasted, Armand Rousseau's Chambertin, Clos de Beze or Clos del la Roche, subjectivity goes out the window. If there is a God, then the only thing he can do for me is, serve them up to me in paradise, then bugger off! I wouldn't share a drop.

However, Pomerol's come a close second, Vieux Chateau Certan, La Conseillant, Certan de May and le Evangile and Petrus run them pretty close.

There is a similarity insofar as the over - riding impression one gets is, "Strawberry Vanilla Crème' Brule'e. Even a hint of brown sugar. It is alleged that these Chateau's mature their wines in Hungarian Oak, which imparts a sweeter vanilla flavour than Allier, Troncais or limousin, or for that matter American Oak, which connoisseur's regard as vulgar.

Chateau's Lafitte, Margaux, Palmer, Haut Brion, Cheval Blanc - although it is pure hedonist indulgence.

Of course, this all subjective. However, if anyone cares to buy a few bottles, I would be more than happy to debate the issue over a glass or two.